Determining Your Pipeline

Brian Genna -

When capturing a facial performance for use with Faceware, there are 4 main ways to go about it. Below is a chart that takes a look at the different types of capture and breaks down what is involved:

Types of Capture Headcam Static HD Cam Body Capture Facial Capture Audio Recording
VO/ADR X X   X X
Faceover X X   X  
Performance Capture X   X X  
Full Performance Capture X   X X X

 

VO/ADR

This method of facial capture is concurrent with final audio capture for your project generally at an audio recording facility. When Full Performance Capture is impossible due to actor considerations, budgets, etc. this is the recommended method of capture. A headmounted camera or a static HD camera can be used to capture video of your performer. This is a very quick method for actors, usually less than five minutes from arrival to ready them for shooting. The main benefit of this method is perfect lip sync in the facial animation, and a performance derived from the actor creating the voice. The main drawbacks are a poorly acted or directed performance will equal poor facial animation, inaccurate eyeline data, and incongruity of facial and body animation timing.

Faceover

This method of facial capture is where an actor mimes back to prerecorded dialog, and can be captured virtually anywhere. When any other is impossible due to actor considerations, budgets, etc. this is the only method of capture for use with faceware. A headmounted camera or a static HD camera can be used to capture video of your performer. This is a very quick method for actors, usually less than five minutes from arrival to ready them for shooting. The main benefit of this method is its independence. It can be done at anytime, by any actor, at any stage of production. The main drawbacks are the extra time needed to properly achieve lip sync, poorly acted or directed performance will equal poor facial animation, inaccurate eyeline data, and incongruity of facial and body animation timing.

Performance Capture

This method of facial capture is where an actor has facial data captured with a headcam concurrently with body mocap under two scenarios. The first is a Faceover scenario with the addition of body mocap being added to the prerecorded vocal performance. The second is when body and face are being captured, but audio is considered “scratch” and would need to be ADR’ed at a later time. Either scenario is used when prerecorded dialog exists or when production quality dialog cannot be captured on the mocap stage, usually due to acoustic or actor scheduling reasons. The Faceover with body is the much more common scenario. The second scenario is rare, as voice actors generally will not ADR body actors, but perform with their own vocal timings. It is good to note, the eyelines from Performance Capture can be joined with a facial performance from VO or Faceover capture to create convincingly unified animation. A headmounted camera should be used to capture video of your performer. This is a relatively quick method for actors, as the headcams can be placed on and focused concurrently with suiting up and ROMing. The main benefit of this method is matching body and facial timing creating a more realistic and unified performance, as well as accurate eyeline animation. The main drawbacks are the extra time needed to properly achieve lip sync (if applicable), and poorly acted or directed performance will equal poor facial animation.

Full Performance Capture

This method of facial capture is where an actor has facial data captured with a headcam concurrently with body mocap and audio recording. This can only be captured in a mocap volume in a soundproof stage. This method is hands down the most convincing way to bring actors performances to the digital realm. There is no substitute for seeing actors to play off each other in body, eyeline, and vocal timings. This is the quickest way of bringing “life” to character animation. If dialog needs to be changed later, it is good to note the eyelines from Full Performance Capture can be joined with a facial performance from VO or Faceover capture. A headmounted camera should be used to capture video of your performer. This is a relatively quick method for actors, as the headcams can be placed on and focused concurrently with suiting up and ROMing. The main benefit of this method is matching body, vocal, and facial timing creating a truly realistic and unified performance, as well as accurate eyeline animation. The main drawbacks are adapting audio pipelines (location sound recordists are a different breed than VO engineers) and poorly acted or directed performance will equal poor facial animation.

Have more questions? Submit a request

Comments

this.location="http://support.facewaretech.com"
Powered by Zendesk