17-03-2014, 10:41 PM
ABSTRACT
An essential form of communication between the director and the
animators early in the animation pipeline is rough cut at the motion
(a blocked-in animation). This version of the character’s performance
allows the director and animators to discuss how the character
will play his/her role in each scene. However, blocked-in animation
is also quite time consuming to construct, with short scenes
requiring many hours of preparation between presentations. In this
paper, we present a puppeteering interface for creating blocked-in
motion for characters and various simulation effects more quickly
than is possible in a keyframing interface. The animator manipulates
one of a set of tracked objects in a motion capture system to
control a few degrees of freedom of the character on each take. We
explore the design space for the 3D puppeteering interface with a
set of seven professional animators using a “think-aloud” protocol.
We present a number of animations that they created and compare
the time required to create similar animations in our 3D user interface
and a commercial keyframing interface.
Index Terms: I.3.6 [Computer Graphics]: Methodology and
Techniques—Interaction techniques; I.3.7 [Computer Graphics]:
Three-Dimensional Graphics and Realism—Animation
1 INTRODUCTION
The heart of a computer animation is the performance of the characters
as they play their part in the narrative arc. Through their motions,
they tell the story, express their emotions and motivation, and
convey their personality. This performance is designed in a process
called blocking-in: working from audio and the storyboards,
the animator creates a rough version of the motion.
Animators often spend one or more days to create a 10-second
blocked-in animation with keyframing in a mouse-based animation
system. Particularly in commercial productions, the blocked-in animation
serves as a form of visual communication between the animator
and the director, and the animator refines and redoes the
blocked-in motion until the director is satisfied that it conveys the
essence of the performance. With several days required between
iterations, this process can be slow to converge. On the other hand,
!takaakis[at]microsoft.com
†{mo, jkh}[at]disneyresearch.com
‡warren.trezevant[at]autodesk.com
for live action movies, a director can guide the performance of actors,
and the actors can rehearse their performance in real time. A
user interface that would allow for such a quick revision cycle for
animation productions would not only reduce the cost of the productions
but might also result in a higher quality product by allowing
more iterations of refinement.
In this paper, we describe a 3D puppeteering interface that allows
an animator to “perform” the motion of a character or a simulation
effect for the blocking phase rather than “keyframing” it in one of
the commercially available animation software packages (Figure 1).
We hypothesize that the animator can perform motion much more
quickly through puppeteering than he/she can animate in traditional
keyframing software. Performing the motion allows the animator to
quickly get a feeling for the timing of the performance while maintaining
approximate control of the poses. A number of animators
have observed that the timing of actions is the key element in conveying
the performance of the character [10, 19].
In addition to keyframing, alternative approaches used to convey
the proposed performance include motion capture, selecting motion
from a database, or videotaping the animator’s performance. Our
preliminary interviews with animators indicated that puppeteering
showed significant promise for several reasons. It allows motions
that are not physically realistic and non-human—a requirement for
many character animations as well as for secondary animations
such as fluid effects. Puppeteering provides the animator with full
control over the design and timing of the motion whereas selecting
from even a large database is necessarily more restrictive. Creating
the blocked-in motion for the actual character geometry and rig
brings it far closer to the final animation than a video of a human
performance or motion capture.
We chose to use an iterative refinement process with a set of animators
to better understand what they expected and needed from
a puppeteering interface for blocked-in motion. Iterative refinement
is the seminal idea of iterating between an interface refinement
phase and a user test phase and has been shown to result in
significantly better systems in many domains. Seven professionally
trained animators and two experienced animators created a variety
of animations with our system. We explored a number of key dimensions
for the interface design with seven animators:
An essential form of communication between the director and the
animators early in the animation pipeline is rough cut at the motion
(a blocked-in animation). This version of the character’s performance
allows the director and animators to discuss how the character
will play his/her role in each scene. However, blocked-in animation
is also quite time consuming to construct, with short scenes
requiring many hours of preparation between presentations. In this
paper, we present a puppeteering interface for creating blocked-in
motion for characters and various simulation effects more quickly
than is possible in a keyframing interface. The animator manipulates
one of a set of tracked objects in a motion capture system to
control a few degrees of freedom of the character on each take. We
explore the design space for the 3D puppeteering interface with a
set of seven professional animators using a “think-aloud” protocol.
We present a number of animations that they created and compare
the time required to create similar animations in our 3D user interface
and a commercial keyframing interface.
Index Terms: I.3.6 [Computer Graphics]: Methodology and
Techniques—Interaction techniques; I.3.7 [Computer Graphics]:
Three-Dimensional Graphics and Realism—Animation
1 INTRODUCTION
The heart of a computer animation is the performance of the characters
as they play their part in the narrative arc. Through their motions,
they tell the story, express their emotions and motivation, and
convey their personality. This performance is designed in a process
called blocking-in: working from audio and the storyboards,
the animator creates a rough version of the motion.
Animators often spend one or more days to create a 10-second
blocked-in animation with keyframing in a mouse-based animation
system. Particularly in commercial productions, the blocked-in animation
serves as a form of visual communication between the animator
and the director, and the animator refines and redoes the
blocked-in motion until the director is satisfied that it conveys the
essence of the performance. With several days required between
iterations, this process can be slow to converge. On the other hand,
!takaakis[at]microsoft.com
†{mo, jkh}[at]disneyresearch.com
‡warren.trezevant[at]autodesk.com
for live action movies, a director can guide the performance of actors,
and the actors can rehearse their performance in real time. A
user interface that would allow for such a quick revision cycle for
animation productions would not only reduce the cost of the productions
but might also result in a higher quality product by allowing
more iterations of refinement.
In this paper, we describe a 3D puppeteering interface that allows
an animator to “perform” the motion of a character or a simulation
effect for the blocking phase rather than “keyframing” it in one of
the commercially available animation software packages (Figure 1).
We hypothesize that the animator can perform motion much more
quickly through puppeteering than he/she can animate in traditional
keyframing software. Performing the motion allows the animator to
quickly get a feeling for the timing of the performance while maintaining
approximate control of the poses. A number of animators
have observed that the timing of actions is the key element in conveying
the performance of the character [10, 19].
In addition to keyframing, alternative approaches used to convey
the proposed performance include motion capture, selecting motion
from a database, or videotaping the animator’s performance. Our
preliminary interviews with animators indicated that puppeteering
showed significant promise for several reasons. It allows motions
that are not physically realistic and non-human—a requirement for
many character animations as well as for secondary animations
such as fluid effects. Puppeteering provides the animator with full
control over the design and timing of the motion whereas selecting
from even a large database is necessarily more restrictive. Creating
the blocked-in motion for the actual character geometry and rig
brings it far closer to the final animation than a video of a human
performance or motion capture.
We chose to use an iterative refinement process with a set of animators
to better understand what they expected and needed from
a puppeteering interface for blocked-in motion. Iterative refinement
is the seminal idea of iterating between an interface refinement
phase and a user test phase and has been shown to result in
significantly better systems in many domains. Seven professionally
trained animators and two experienced animators created a variety
of animations with our system. We explored a number of key dimensions
for the interface design with seven animators: