Motion capture

• 1. Introduction

  ›

  Motion capture, the process of recording high-resolution human and object movement into a computer system, began not in Hollywood but in biomechanics labs. Researchers in the 1970s and 1980s were trying to understand how people move. The technology they built to measure joint angles and limb trajectories eventually found its way into every corner of entertainment, medicine, robotics, and sport.

  The key question that motion capture answers is deceptively simple: how do you make a digital character move like a real person? The answer turns out to require cameras, sensors, reflective balls, and sometimes LEDs communicating at up to 960 frames per second. What started as a scientific measurement tool became the engine behind some of the most recognizable characters in cinema history.

  The first virtual actor fully animated by motion capture arrived in 1993, produced by Didier Pourcel and his team at Gribouille. They cloned the body and face of French comedian Richard Bohringer using what Pourcel himself described as still-nascent tools. That moment planted a flag. The decades that followed would see motion capture spread from arcade games to Olympic athletics research, from underwater swimming pools to the world's largest indoor drone testing facility in Indiana.

• 2. From Labs to Arcades

  ›

  As early as 1988, motion capture was already being used in commercial video games. Martech's game Vixen used it to animate 2D player characters, with the movements performed by model Corinne Russell. That same year, Magical Company's arcade fighting game Last Apostle Puppet Show used the technology to animate digitized sprites. Neither was cinema-scale, but both signaled that the technique had commercial appeal beyond the research lab.

  Sega pushed the technology further with its Model arcade series. Virtua Fighter in 1993 and Virtua Fighter 2 in 1994 used motion capture to drive 3D character models in a way that felt weightier and more physical than hand-keyed animation could easily achieve. Namco's 1995 arcade game Soul Edge went a step further, deploying passive optical system markers for its capture work.

  By mid-1995, developer and publisher Acclaim Entertainment had built its own in-house motion capture studio directly into its headquarters. That investment had consequences beyond games. Warner Bros. later acquired Acclaim's motion capture technology specifically to use in Batman Forever (1995), and Acclaim's own video game tie-in for that film used the same captured data to animate its digitized sprite graphics.

• 3. Hollywood's Adoption

  ›

  The 1990 film Total Recall contains a brief but historically significant motion capture sequence: the scene in which Arnold Schwarzenegger and other actors walk through an X-ray machine was shot using the technique. It was a supporting effect, not a starring role.

  The 1999 film Star Wars: Episode I - The Phantom Menace changed the scale of ambition. Jar Jar Binks, played by Ahmed Best, was the first main character in a feature-length film created using motion capture. A year later, the Indian-American film Sinbad: Beyond the Veil of Mists became the first feature-length film made primarily with the technique, though it had a very limited release. 2001's Final Fantasy: The Spirits Within was the first widely released movie built on motion capture technology. Despite a poor box office, its supporters took notice and the technology kept advancing.

  The Lord of the Rings: The Two Towers introduced real-time motion capture to feature filmmaking. Actor Andy Serkis's movements were streamed directly into the computer-generated skin of Gollum as the performance was being recorded. James Cameron's Avatar pushed this further still. During filming, Cameron used Autodesk MotionBuilder software to render a real-time screen image so that both director and actor could see the result as it was being shot. Cameron was confident enough in his system that he invited Steven Spielberg and George Lucas on set to observe it in action.

• 4. Performance Capture and Its Reach

  ›

  When motion capture expands to include face, fingers, and subtle expressions, the industry calls it performance capture. Mark Ruffalo used this technique in Marvel's The Avengers to play both the human Bruce Banner and the Hulk, making him the first actor to portray both versions of the character in a single film. Earlier Hulk depictions had kept the creature purely computer-generated.

  Of the three nominees for the 2006 Academy Award for Best Animated Feature, two (Monster House and the winner Happy Feet) used motion capture. Only Disney and Pixar's Cars was animated without it. Pixar made a pointed statement in the credits of their next film, Ratatouille, stamping it "100% Genuine Animation - No Motion Capture!" The debate about where performance ends and animation begins was clearly alive enough to warrant a public declaration.

  Robert Zemeckis became a sustained advocate for the technology. The Polar Express used motion capture to allow Tom Hanks to perform as several distinct digital characters, providing the voices for each as well. Disney later acquired Zemeckis's production company ImageMovers Digital in 2007, but closed it in 2011 following the poor box-office performance of Mars Needs Moms. Television found its own uses: series produced entirely with motion capture animation include Laflaque in Canada, Sprookjesboom in the Netherlands, and Headcases in the United Kingdom.

• 5. The Technology in Detail

  ›

  Passive optical systems are built on retroreflective markers. Small rubber balls coated in reflective tape are velcroed onto a full-body spandex or lycra suit, or in biomechanics contexts applied directly to skin. Cameras calibrated to provide overlapping projections then triangulate each marker's position in three dimensions. A typical system uses somewhere between 2 and 48 cameras, though systems of over 300 cameras exist to reduce a problem called marker swapping, which occurs because all passive markers look identical. These systems typically run at 120 to 160 frames per second, though by narrowing the region of interest they can reach 10,000 frames per second.

  Active marker systems solve the identification problem differently. Each LED marker emits its own light rather than reflecting it, which increases the range and volume available for capture. Measurement resolution can reach down to 0.1 mm within the calibrated volume. One eight-camera, 12-megapixel system running at 120 hertz costs around $20,000. The TV series Stargate SG-1 used an active optical system for its visual effects work, as did Industrial Light and Magic on Van Helsing, where active markers allowed capture of Dracula's flying brides across very large sets.

  Inertial systems take a different approach entirely. They use miniature sensors containing a gyroscope, magnetometer, and accelerometer, transmitting data wirelessly to a computer. No external cameras are needed for relative motion, which means capture can happen in tight spaces or across large outdoor areas. Suits using inertial systems range from around $1,000 to $80,000. The tradeoff is positional drift, which accumulates over time. Mechanical exoskeleton systems avoid this by directly tracking joint angles through rigid structures of jointed metal or plastic rods linked at body joints, typically priced between $25,000 and $75,000.

• 6. Science, Sport, and Medicine

  ›

  Purdue University houses the world's largest indoor motion capture system inside its Purdue UAS Research and Test facility, known as PURT. The facility uses 60 motion capture cameras to cover a volume of 600,000 cubic feet, tracking targets with millimeter accuracy. Its purpose is drone research. Airspace regulations limit outdoor experiments with unmanned aerial systems, so indoor volumes like PURT allow researchers to run tests that would otherwise be legally or practically impossible.

  In clinical medicine, gait analysis uses motion capture to evaluate human movement across multiple biomechanical factors, with results streamed live into analytical software. Pose detection extends this into rehabilitation, providing continuous monitoring and individually tailored programs for patients recovering from surgery or injury. Physical therapy clinics have adopted the technique as an objective way to track patient progress.

  In 2020, the two-time Olympic figure skating champion Yuzuru Hanyu completed his thesis at Waseda University. Using data from 31 sensors placed on his body, he analyzed his jump mechanics and evaluated how motion capture technology could improve both the scoring system and the training of skaters. A summary of the thesis was published in an academic journal in March 2021. The same tools that Gribouille used in 1993 to clone Richard Bohringer had, across two decades, become precise enough to study the physics of an Olympic jump in academic detail.