Motion Capture

Motion Capture

Capture and process motion capture data for realistic character animation in games and film.

Overview

This skill covers the complete motion capture pipeline from session planning to final animation. Learn marker placement, data capture techniques, cleaning and solving mocap data, retargeting to custom characters, facial mocap workflows, and integrating mocap into production pipelines.

Quick Start: Mocap System Selection

Core Mocap Concepts

Motion Capture Types

Optical Systems:

• Infrared cameras track reflective markers
• High accuracy (sub-millimeter)
• Requires controlled environment
• Industry standard for film/AAA

Inertial Systems:

• Wearable sensors (IMUs)
• No cameras required
• Works anywhere (outdoor, on-set)
• Slightly less accurate than optical

Markerless Systems:

• Computer vision, no markers/suits
• Webcam or video-based
• Lower accuracy but very accessible
• Improving rapidly with AI

Hybrid Systems:

• Combine optical + inertial
• Best accuracy and flexibility
• Used in high-end productions

Mocap Pipeline Stages

Stage 1: Pre-Production Planning

Define Requirements:

• What actions/movements needed?
• How many performers?
• Props and environment needs?
• Quality requirements?
• Budget and timeline?

Performer Selection:

• Match character proportions if possible
• Skilled performers vs. actors
• Stunt performers for action
• Facial capture considerations

Session Planning:

• Shot list and priorities
• Timing and scheduling
• Backup plans
• Reference video

Stage 2: Calibration and Setup

System Calibration:

• Camera calibration (optical)
• Sensor calibration (inertial)
• Volume calibration (capture space)
• Verify accuracy

Marker Placement:

• Follow standard marker sets (e.g., Vicon Plug-in Gait)
• Anatomical landmarks
• Rigid body markers for props
• Minimize occlusion

Test Capture:

• Capture test movements
• Verify marker visibility
• Check data quality
• Adjust as needed

Stage 3: Capture Session

Recording:

• Slate each take (identify in data)
• Capture reference video
• Multiple takes for safety
• Monitor data quality in real-time

Direction:

• Communicate clearly with performer
• Review takes immediately
• Iterate and refine
• Capture variations

Stage 4: Data Processing

Cleaning:

• Fill gaps (missing marker data)
• Remove noise and jitter
• Fix marker swaps
• Smooth trajectories

Solving:

• Reconstruct skeleton from markers
• Solve for joint rotations
• Handle occlusions
• Verify solve quality

Export:

• Export to FBX, BVH, or other formats
• Include skeleton and animation
• Proper scale and orientation
• Metadata and notes

Stage 5: Retargeting

Character Mapping:

• Map mocap skeleton to character skeleton
• Handle proportion differences
• Adjust for different bone counts
• Maintain motion quality

Refinement:

• Adjust foot contact
• Fix penetrations
• Polish hand poses
• Add secondary animation

Software and Tools

Capture Software

Processing Software

• MotionBuilder: Industry standard for mocap cleanup and retargeting
• Maya: Full pipeline, HumanIK for retargeting
• Blender: Free option, improving mocap tools
• iClone: Real-time character animation with mocap

Cleanup Tools

• Shogun Post: Vicon's post-processing software
• MotionBuilder: Excellent cleanup and filtering tools
• Custom scripts: Python/MEL for batch processing

Best Practices

Marker Placement

1. Anatomical accuracy — Place markers on bony landmarks
2. Minimize occlusion — Avoid markers blocking each other
3. Rigid clusters — Use 3+ markers for rigid bodies
4. Symmetry — Mirror left and right sides
5. Documentation — Photo reference of marker placement

Capture Session

1. Calibrate frequently — Recalibrate if setup changes
2. Capture ROM — Range of motion for each performer
3. Reference video — Always record video reference
4. Multiple takes — Capture backups of important actions
5. Real-time review — Check data quality during session

Data Cleanup

1. Work non-destructively — Keep original data
2. Fill gaps carefully — Use appropriate interpolation
3. Smooth judiciously — Don't over-smooth and lose detail
4. Verify solve — Check skeleton solve quality
5. Document issues — Note problems for future reference

Common Challenges and Solutions

Challenge: Marker Occlusion

Solution: Use more cameras from different angles, adjust marker placement, use predictive gap filling, capture multiple takes.

Challenge: Foot Sliding

Solution: Lock feet during contact, adjust retargeting, use IK constraints, manually key foot positions.

Challenge: Proportion Mismatch

Solution: Use proper retargeting tools, adjust limb lengths, use IK for extremities, accept some differences.

Challenge: Noisy Data

Solution: Apply appropriate filtering, check calibration, improve lighting (optical), reduce magnetic interference (inertial).

Using the Reference Files

When to Read Each Reference

/references/optical-systems.md — Read when setting up optical mocap systems, calibrating cameras, placing markers, or troubleshooting optical capture issues.

/references/inertial-systems.md — Read when using inertial mocap suits, calibrating sensors, capturing on-location, or solving inertial data.

/references/markerless-systems.md — Read when using markerless/AI-based mocap, processing video for mocap, or working with accessible mocap solutions.

/references/facial-mocap.md — Read when capturing facial performances, setting up facial rigs for mocap, processing facial data, or integrating facial animation.