MoCap options abound

August 2012, The exhibit floor at Siggraph 2012 featured a number of difference motion capture (MoCap) options for creating CG and digital duplicate characters of live actors. These range from passive suits to active MEMS technology and a number of options in between. None of these technologies is a do-all end-all, rather, for full content creation, a number of these systems need to be used on the same project.

In the panel that was discussing the film “the Avengers”, the main motion capture system was a traditional high resolution optical system that used passive tracking. In this mode the actors wore a mocap outfit that had painted dots on the key movement areas (arms, legs, hands, torso, etc) that are tracked by a 4-6 camera monitoring system. This method was used based on the range of motion of the actors and they fact that they would be performing stunts in the outfit, and it would minimize delays due to damage to the sensors if they were electronic and had impacts during the stunts.

For more traditional green screen capture and for game development, finer granularity in movement is required than standard passive tag locations. These applications, active sensors that include RF, UV, or IR light sources and detectors rather than standard visible light cameras are used. This allow for movements with twisting and other movements that have optical light of sight blockages to still be captured. These movements are all real time capture and the time grids for the capture are only at discrete sampling points in the system.

MEMS based systems not only can be distributed over larger areas, but they provide more data for the movement. Instead of a simple time based motion vector, velocity, acceleration and elevation in addition to both relative and absolute position can be recorded. Whereas most optical systems record in the 30-60hz level, some MEMS systems operate up to 200Hz giving a much more fluid response curve. The MEMS systems however, have a significant cost delta from the passive systems, are prone to damage from impacts and water and have limited operating cycle times to the requirement to be remote powered.

Finally, for facial and detailed body movements, high resolution passive systems with close located cameras are used. These allow for the mapping of eye movement, lip and nose movement and full facial expression details to be captured. These techniques were used for Planet of the Apes and Avatar to produce the digital doubles for the actors.

All of these systems integrate with the motion capture tools and motion graphics software. They are a key driver for multi-frame rendering. The major tools all integrate with MoCap data from the big providers. As there is no data standard in place, systems from some of the smaller vendors while capturing quickly and at high resolution, cannot easily be integrated into the CG flows as they do not connect with the software APIs for their tools.

The realism required from CG effects is pushing MoCap to be a mainstay of feature films and it is now progressing to be a mainstay feature, rather than a luxury, for TV production. The ability to create and apply realistic movements to a CG character rather than have costly makeup and FX for the actor brings both creativity and cost reduction to the creative design projects.