Editing and constraining kinematic approximations of dynamic motion

  • Authors:
  • Cyrus Rahgoshay;Amir H. Rabbani;Karan Singh;Paul G. Kry

  • Affiliations:
  • McGill University;McGill University;University of Toronto;McGill University

  • Venue:
  • I3D '12 Proceedings of the ACM SIGGRAPH Symposium on Interactive 3D Graphics and Games
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

Physical simulation is now a robust and common approach to recreating reality in virtual worlds and is almost universally used in the animation of natural phenomena, ballistic objects, and character accessories like clothing and hair. Despite these great strides, the animation of primary characters continues to be dominated by the kinematic techniques of motion capture and above all traditional keyframing. Two aspects of a primary character in particular, skeletal and facial motion, are often laboriously animated using kinematics. There are perhaps three chief reasons for this. First, kinematics, unencumbered by physics, provides the finest level of control necessary for animators to breathe life and personality into their characters. Second, this control is direct and history-free, in that the authored state of the character, set at any point in time is precisely observed upon playback and its impact on the animation is localized to a neighborhood around that time. Third, animator interaction with the time line is WYSIWYG (what-you-see-is-what-you-get), allowing them to scrub to various points in time and observe the character state without having to playback the entire animation. Secondary dynamics can be overlaid on primarily kinematic character motion to enhance the visceral feel of their characters. But unfortunately compromise the second and third reasons animators rely on pure kinematic control.