What Is Motion Capture in Animation?

Motion capture is the process of recording real people moving and putting those actions into computer figures. 

In fact, animators don’t have to carefully make each frame by hand; rather, actors wear suits with sensors or marks on them, and computers track their moves to make cartoons that look real. 

For example, real actors played the parts of Gollum in The Lord of the Rings and Thanos in Avengers, and technology then moved their moves, reactions, and feelings to CGI characters.

Motion Capture Meaning

Motion capture is a technique used in animation to record the movements of real people using sensors or cameras. This information is then sent to computer figures so they can move naturally. 

Here’s how it works. An actor performs while wearing a special suit with sensors, and a computer copies every move they make onto a 3D animation. It looks like the figure moves just like the real person.

Know the Lord of the Rings character Gollum? We call that motion capture. Andy Serkis played out every move Gollum made, and artists then put it on the computer-generated character. Similarly, Josh Brolin wore a mocap suit to play Thanos in Avengers.

“Mocap” or “mo-cap” is another name for it, though. Also, the term “performance capture” is used interchangeably and means recording more than just body movement, such as the actor’s face and feelings.

For example, do you remember Avatar? Real actors practiced those blue aliens’ moves first, which is why they looked so real. Everything, like walking, running, and facial emotions, was recorded by the technology and then put on the Na’vi figures.

3 Types of Motion Capture Systems in Animation

The mocap method comes in three different types, and each one feels different.

1. Marker-Based Tracking (Optical Tracking Systems)

Have you seen those behind-the-scenes videos where the players are dressed in black clothes with white ping-pong balls stuck all over them? 

That’s tracking with markers, and it means that 20 to 50 bright marks are put on the actor’s shoulders, elbows, wrists, knees, and other places on their body that bend or move.

How It Works:

1. Each camera sends out infrared light that hits those marks and makes them flash.
2. Based on the view of each camera, the marks light up like tiny signals.
3. The computer can figure out exactly where each sign is in 3D space because it sees them from different cameras.
4. This calculation is done 120 to 160 times every second.
5. When all of those marks are put together, the system builds a computer frame that perfectly copies your actions.

The multiple-angle view is what makes this system so exact; even if your arm stops the view of a sign from one camera, it can still be seen clearly by five other cameras.

2. Markerless Motion Capture

Here, you don’t need a suit, and you don’t have to stick marks on it. The actor wears regular clothes and moves around while cameras record. After that, smart AI software watches the video to track the actor’s movements.

How It Works:

1. Usually, one or more standard cameras record the person moving. 
2. Then, AI programs look through each animation frame and try to find the shape of a person among all the other things in the shot.
3. The software then knows what a head looks like, what arms look like, and how legs bend because it learned from training by watching thousands of movies. 
4. It keeps track of where each body part moves frame by frame.
5. In the end, the 3D skeleton gets constructed purely from this visual analysis.

3. Inertial Motion Capture

Inertial capture is when you wear a suit full of tiny motion sensors instead of cameras that watch you. 

They use IMUs, or Inertial Measurement Units, and these IMUs are small electronic boxes that are pressed right into the fabric. 

Also, there are usually 17 to 19 sensors spread across your body at every major joint and these sensors can measure their own movement without the need for cameras to point at them.

How It Works:

• Gyroscopes inside measure how fast the sensor spins.
• Accelerometers can tell when speed and direction change.
• All of the sensors record their movement 60 to 240 times per second at the same time.
• The data sends itself wirelessly to a nearby computer using Wi-Fi or Bluetooth.
• Software puts together all the readings from the sensors to recreate your full body movement based on how each joint moved.

How Motion Capture Works?

Here are the steps you need to take to complete the process.

Step 1: Getting the Actor Ready (Dressing Up)

After getting ready, the star puts on a cool suit. Sometimes it’s a tight black outfit that fits close to the body, like a wetsuit, but made of fabric that can stretch.

Later, the team covers the suit with sensors or marks.

• Shoulders, elbows, wrists
• Hips, knees, ankles
• Spine, chest, head

If they need to make precise hand moves, they sometimes use their fingers.

The number of marks used varies from 15 to 50, based on how thorough the catch needs to be. 

Generally, 15 to 20 marks are enough for a simple body capture, but for something more complicated, like a martial arts fight scene, they might need 40 or more to catch every little move.

It’s either tiny LED lights that glow on their own or shiny dots that shine when light hits them.

The actor also wears a hat rig with a small camera fixed right in front of their face, and this setup records every little movement of the face, like when someone raises an eyelid or talks, it records it.

Step 2: Setting Up the Capture Space

Special cameras are placed all around the room, usually between 8 and 48 in a circle or square design to cover every angle. 

The infrared sensors in these cameras can see those shiny marks even when the room isn’t well-lit.

Anytime the character moves around a lot, the recording area needs to be pretty big, so it’s sometimes a 20-foot-by-20-foot room.

Step 3: Calibration (Checking to See if Everything Is Right)

They have to set the system before they can start recording, and it means showing the computers and cameras exactly where everything is in the room.

A helper usually moves around the capture area with a specific calibration device, which is often simply a stick with marks on it at established places. 

Then, the system uses all of the cameras to look at this item and figure out the exact 3D plan of the room.

The character typically stands straight up with their arms out to the sides in a T-pose, and the system keeps track of this as the starting place.

Step 4: Recording the Performance (The Actual Capture)

Finally, the player acts. They move around the room and do whatever the scene calls for, like climbing, fighting, dancing, walking, or just talking.

Here’s what happens at the same time as the character animation moves:

• The cameras generally take pictures at 120 to 160 frames per second.
• Each camera gets a different view of the marks and sends that information to the computer.
• The computer generates a digital skeleton that moves in perfect time with the performer by comparing the images from all the cameras.
• After that, the director uses a screen to keep an eye on everything.

In most mocap sessions, the same action is done more than once to get different results or to get a perfect recording of it. 

So that the director can pick the best one, a character might walk around the same set of steps 5 or 6  times.

Step 5: Clearing up the Raw Data

Some cleaning up is generally needed on the raw mocap data after it has been recorded. The computer has stored thousands of data points, which are the positions of each marker every fraction of a second. However, there are glitches and mistakes from time to time.

Common problems that need to be fixed:

• “Marker swap” happens when two marks are too close to each other for the computer to tell them apart.
• Occlusion is when the actor’s body briefly blocks the camera’s view of a sign.
• Noise is fluttering, random moves that aren’t part of the show

A technical artist goes through the data frame by frame and fixes these mistakes and fills in blanks where marks couldn’t be seen.

Step 6: Retargeting the Motion (Putting It on the Character)

Retargeting is the process of moving motion data from one body to another.

More interestingly, the actor’s body features are clearly different from those of the CGI figure. 

For example, the character could have longer arms or shorter legs, or it could be a creature that is built very differently from humans.

The software that does the retargeting has to change the moves so that they fit the new body while still making the performance look natural.

Step 7: Adding the Final Touches (Secondary Animation)

Mocap does a great job of capturing the main moves of the body, but it can’t catch everything. So details like hair moving, clothes moving, or a cape blowing in the wind need to be added later.

Afterwards, animators add the following details by using physics modeling and hand animation:

• Hair dynamics (how hair bounces and sways with movement)
• Cloth simulation (how fabric wrinkles and flows)
• Muscle jiggle and skin deformation
• Any over-the-top moves that aren’t naturally human

What Is Motion Capture Used For?

Motion capture is used whenever real human action needs to be recorded, studied, or recreated. It was first used in movies to bring CGI characters to life. 

Now it’s used to train doctors and soldiers, help players improve their game, make video games feel real, and even help injured people get better. Some other uses of motion capture are:

• Creating realistic CGI characters (Gollum, Thanos, Avatar’s Na’vi)
• Animating creatures and monsters
• Planning complex action scenes during pre-production
• De-aging actors or creating younger versions

• Animating player characters with realistic movements
• Recording athlete motions for sports games (FIFA uses real soccer players)
• Capturing fight choreography for action games

What is a Motion Capture Suit?

A mocap suit is a tight-fitting black outfit with sensors or reflective markers all over it that follow your movements. The suit generally consists of stretchy spandex or Lycra, which is the same material that dancers and athletes wear. Some types of suits and equipment include:

• Basic optical suit: Black spandex outfit with 15-50 reflective marker dots attached at joints
• Inertial suit: Contains built-in wireless sensors (IMUs) that don’t need cameras
• Full-body suit: Covers the entire body, including hands and sometimes individual fingers
• Facial 3D rigging: Helmet with a mounted camera pointing at the face to capture expressions
• Gloves with markers: Special gloves for detailed finger and hand tracking
• Foot trackers: Sensors on shoes or feet for precise ground contact data

Standard sizes for the suits are S, M, L, and XL, and they can cost anywhere from $1,000 for a simple setup to $20,000 or more for a professional system with all the sensors. 

Must-use Motion Capture Software

Popular motion capture software for game and animation includes Autodesk MotionBuilder for handling data, Vicon Blade for optical systems, and Rokoko Studio for inertial capture.  

What Are the Best Motion Capture Animation Movies?

These are the best mocap movies: 

• Avatar (2009)
• The Lord of the Rings trilogy (2001–2003) 
• Planet of the Apes series (2011–2017)
• Avengers: Infinity War and Endgame (2018–2019)
• Alita: Battle Angel (2019)

Though almost every big hit movie these days uses mocap in some way, from the new Star Wars movie to the version of “The Lion King” to “Ready Player One.”

When Was Motion Capture First Used in Video Games?

Motion capture first appeared in video games in the late 1980s, a lot earlier than most people think. 

In the first real example, a game called Vixen from 1988 used model Corinne Russell’s moves to create 2D character sprites. However, the technology back then was very basic compared to what we have now.

Plus, Prince of Persia, which came out in 1989, was the big breakthrough, and many people think it was the first real motion capture game. The person who made the game filmed his brother doing parkour moves and running around. 

Then, he copied those actions frame by frame to make the characters’ realistic motions. This is a lot like rotoscoping and early mocap methods put together. 

Milestones in gaming mocap are:

• 1988: Vixen used early mocap for 2D sprite animation
• 1989: Prince of Persia traced real human parkour movements
• 1993: Virtua Fighter became the first 3D game using proper mocap technology
• 1994: Virtua Fighter 2 improved the system significantly
• 1995: Acclaim Entertainment built an in-house mocap animation studio for game development
• Late 1990s: Games like Crash Bandicoot and Spyro the Dragon started using mocap regularly
• 2000s onward: Mocap became standard for realistic games, especially sports titles like FIFA and NBA 2K

What Is the Difference Between CGI and Motion Capture?

CGI stands for “Computer-Generated Imagery,” and it’s a broad term for everything that was made digitally on a computer. Well, it includes characters, settings, explosions, and just about any other visual effect that wasn’t captured on film. 

On the other hand, motion capture is a way to make CGI figures by recording the moves of real people and adding them to digital models. Therefore, mocap is a tool that helps make CGI and not something different from it.

Conclusion | Motion Capture in Animation

Motion capture has totally changed how we make digital figures. It’s now possible to turn real human acts into beautiful cartoons, which wasn’t possible even 20 years ago. The technology keeps getting better and easier to get. 

For example, marker-based systems record every small movement in Hollywood movies, markerless AI tracking is available for independent makers, and portable inertial suits can be used anywhere outside. 

Since Gollum’s big step forward in 2002 to today’s real-time virtual production stages, mocap has been the link between digital thought and human feeling.