Two skeletons, half the work: the smart tech hidden in our software

Written by

14 May 2021

You’ve heard about Cortex – our most powerful software yet, which offers a complete set of tools for motion tracking and editing – and you’ve heard about BaSix Go, our new, easy-to-use software with a setup time that takes under a minute. 

But do you know about the groundbreaking hidden tech inside our software? 

For decades, animators using mocap data have had to find methods to convert the 3D-tracked position of markers into the motion of a humanoid skeleton rig. Previous methods included using vector algebra – which was more suited to biomechanists who needed rigorous, repeatable, and reproducible data as well as a method that could calculate joint-centred angles – and global optimisation, which is the method most animators will want to use for motion capture as it allows them to import a Skeleton with fixed length bones that can be scaled up and down to fit the bone length needed for the character they want to animate. 

IKendo skeleton

 

But Motion Analysis has developed a third way to capture human motion. We started asking the question: What if there were two Skeletons? 

And so Ikendo was born. 

Building on Calcium Solver’s global optimisation method, Ikendo makes use of Motion Analysis’s AMRs (Active Marker Rigs) and can be used with both our classic Cortex software or our more affordable BaSix system. It consists of two Skeletons: one constrained by markers to match the mocap subject, and one that matches the animator’s rig. 

Meet the Ikendo Skeleton 

The first Skeleton has well-defined feet, head, hands and a root segment (usually the pelvis), and the markers keep accurate track of those segments. Because of the immense precision of the tracking we can then calculate the intermediate segmental positions. For example, we would know the exact position of the pelvis, and the exact position of a foot, but we can use further information – such as the knee being a hinge, the hip a ball joint and the ankle a universal joint  – to calculate where the thigh and shank must be to fit the data. This means we can calculate all of those joint centres – not just create a missing marker, but actually calculate all of those segments –  for a very well-defined, specific Skeleton.

We then import the user’s Skeleton and use virtual markers that are created with our internal Skeleton to act as the “springs” which drive their Skeleton.

 

The secret to BaSix Go’s quick setup time

Our BaSix Go software has always been powered by Ikendo, we’ve just finally put a name to the smart tech at the core of this system. 

By separating the retargeting from the inverse kinematics of the tracked skeleton, marker placement is simplified, and this enables us to produce our infamous one-minute mocap setup. 

We have programmed the character in the Ikendo Skeleton into a T-pose, which basically allows the system to compute how long your arms and legs are, enabling us to then scale the skeleton. Of course, this endoskeleton has to be scaled to agree with the subject being tracked, but all that requires is a simple click of the mouse. 

AMR markers are automatically identified out of the box as soon as they are switched on. The mocap performer simply needs to attach the six AMRs onto their hands, feet, waist and head, stand in a T-pose and we are able to sync. Nothing more is needed for the tracking.

BaSix set up is quick and easy

 

Ikendo not only makes the software simpler and quicker to use, but it lowers the overall cost of using the mocap tech and is compatible with all Motion Analysis streaming partners. 

Want to find out more about our systems, and how Ikendo could make your mocap experience much simpler? 

Book a demo with us here