Motion capture vs Performance Capture
There is a lot of confusion surrounding the process of motion capture. How much of the production is a result of the actor versus the role of the post-production team? Is a studio needed to perform motion capture? Is mocap only used in big-budget productions? All of these questions and more result in a lot of misconceptions around mocap. While a lot of people may be aware of the idea of mocap thanks to Hollywood blockbuster productions, fewer may have heard of the term ‘performance capture’, and even fewer would be able to name the differences between ‘performance’ and ‘motion’ capture.
Motion capture
As we explained in , ‘Motion-Capture 101’, mocap can be described as the process of translating an actor’s movement through a mocap suit into data and then placing this onto a virtual skeleton through software such as GIANT, which mirrors the actor’s movements. Once the desired performance is achieved, the post-production team then animates the digital character on top of the virtual skeleton within a program such as MotionBuilder. This allows the team to ‘clean up’ the motion capture data, and refine the movements of the character for the final version.
A notable example of ‘motion capture’ would be the performance of Andy Serkis as Gollum in The Lord of the Rings: The Two Towers. That particular performance was captured in three stages. First, Andy Serkis, performed as Gollum, sans mocap suit, on set with other actors. He then went to a mocap studio stage, which is enfitted with numerous cameras used to capture an actor’s movements from all angles. He then recreated his on-set movements while ‘puppeteering’ a digital Gollum. Following this, the process moved to post-production, where Gollum’s face, hands, feet, and facial expressions were animated using Andy Serkis’ performance as a reference.
Performance Capture
What separates ‘performance capture’ from the creation of a character such as Gollum, is that, while an actor’s movements are usually recorded in the same way, there is also the introduction of facial capture. Accurately capturing an actor’s facial expressions has often been a sticking point for motion capture productions. While body motion capture traditionally only needs a smaller set of markers to record an actor’s movements with a high level of accuracy, a larger amount of markers in a much smaller size would be needed to capture the nuances of facial expressions. As well as being practically more challenging for an actor, this also makes the data solving and retargeting process much more challenging for the team in post.
To solve this problem, many modern studios, including Animatrik, have turned to camera-based solutions, whether this is through head-mounted cameras or seated capture cameras. Rather than marker tracking, this solution is based on a 3D scanning method which involves translating a recorded performance that takes every frame of video recorded performance and digitizing that into 3D form. What this means for the team in post-production is that there is less data to re-map, and if retargeting is needed, they have the highest data set to work with.
For an actor, this means much more of their performance is being translated onto a digital character as opposed to a strictly motion-captured performance. For comparison, a more recent Andy Serkis performance would be ‘Snoke’ in 2017’s ‘The Last Jedi’. Working with ILM and Animatrik, Andy Serkis wore a ‘head-mounted camera’ and a mocap suit while acting on-set. What this means is that his face and body motions are captured simultaneously, resulting in a more connected, and nuanced, set of data to be translated onto the digital software.
‘Performance capture’ is exemplified by performances that involve acting with the whole body and face, with some well-known examples including Josh Brolin’s performance as ‘Thanos’ in ‘Avengers: Endgame’ and Andy Serkis as ‘Caesar’ in ‘Planet of the Apes’. In contrast to an early motion-captured performance like ‘Gollum’, these performances are more of a one to one representation of an actor’s time on a mocap stage. Where in the past performances were captured in parts, and replicated by VFX teams digitally, the evolution of performance technology means that the process is smoother for actors and production teams alike.