Definition
Motion capture (mocap) is technology that records the real movements of a person or object with sensors or cameras and turns them into digital data. When you apply that data to a computer-generated character, a virtual person or animal moves as smoothly and naturally as a real one.
The Secret of the Little Balls on the Body
If you've watched a behind-the-scenes movie documentary, you may have seen actors in skin-tight suits covered with small ping-pong-like balls. These little balls are called markers, and they reflect light strongly.
Dozens of special cameras around the studio watch only the markers. They ignore the actor's face and the color of the clothes. They quickly track just the shiny dots in real time. The computer combines views from many angles and calculates the exact position of each marker in 3D space.
Once it knows where the markers are, the computer uses the positions of the many markers on the body's surface to estimate where each joint center is. Then it lines up a premade virtual skeleton to match. When the actor raises an arm or bends a knee, the virtual skeleton moves in real time with the same angle and speed. Put a character's graphics over that skeleton, and you get a lifelike 3D character.
To Be a Bit More Precise
To be a bit more precise, motion capture doesn't always happen in a special studio full of markers and cameras. Depending on the situation, there are many methods that don't use cameras at all.
One common method attaches inertial sensors to clothing, similar to the sensors inside your smartphone or smartwatch. These sensors directly detect the body's rotation and tilt. So it's no problem if something blocks the camera's view, and you can capture movement freely, even outdoors, anywhere you like. The catch is that the sensors keep adding up their readings, so small errors called drift build up over time and the position slowly slips. Nearby metal and magnetic fields can also interfere.
In recent years, computer vision (technology that analyzes images) and artificial intelligence have advanced a lot. Now you don't even need special sensors or a suit with markers. Markerless technology is spreading fast, where AI finds a person's joints on its own in everyday video shot with a regular smartphone camera.
Performance Capture: Recording Expressions and Emotions
Early motion capture could only record big movements of the arms and legs. Today's technology reads far finer details: the tiny tremble of each finger joint, the movement of the eyes, and the subtle twitch of muscles around the lips.
Technology that captures not just body movement but also an actor's facial expressions, voice, and emotional arc all at once is called performance capture. An actor wears a helmet with a tiny camera pointed at the face. Then even a computer-generated alien or a giant monster can inherit the actor's deep emotional performance and come to life.
Today, motion capture goes well beyond film and games. It's widely used in many real-world fields, such as correcting athletes' form, simulator training for flying planes or driving cars, and teaching precise movements to industrial robots.
Common misconceptions
Once you use motion capture, graphic designers have no work left and everything finishes automatically.
Motion capture data is just the coordinates of a skeleton. Careful post-production is still essential: smoothing out jumpy errors in the data and naturally dressing the character with skin and clothes.
Where you meet it
In the movie Avatar, actors wore suits covered in markers and small cameras on their heads to film their faces, which brought the virtual alien characters to life.
In sports video games, a famous soccer player's unique shooting form and dribbling style are applied directly to the game character.
In one sentence
It's technology that collects real human movement data and maps it onto a digital skeleton, giving virtual characters a sense of life.