e-Motion: a system for the development of creative animatics
Santiago Negrete-Yankelevich and Nora Morales-Zaragoza
División de Ciencias de la Comunicación y Diseño
Universidad Autónoma Metropolitana (Cuajimalpa)
Av. Constituyentes 1054, México D.F. 11950, México
{snegrete/nmorales}@correo.cua.uam.mx
Abstract
This paper introduces e-Motion, a software system for
the creation of animatics1
, which are important tools
within the process of creation of animated graphics for
TV. This type of animation, generated by the system
from plots in plain text, allows production teams to
envision how a final motion graphics piece can be developed.
We argue that our system plays a creative
role within the generative process. Specifically, our
work is linked to a real production team, involved in
the creation of animated shorts, called Imaginantes,
for Mexican television.
Introduction
Computer systems intervene more and more in creative
practices. They play different roles in teams of people
working on projects that produce innovative work and
whose overall process can be deemed creative by a suitably
selected group of human experts. Just as in the case of creative
teams formed strictly by human members, the blame
for creativity can be distributed amongst the team members
including computer systems (Maher,M.L. 2012).
In this paper we describe e-Motion, a computer system
that builds animatics for a pre-production process to create
motion graphics. In order to test the system we embed it in
the process of Imaginantes, a TV production of a series of
animated shorts (one minute long) based on texts of different
authors aimed at encouraging viewers to get involved
in Music, Literature, Fine Arts and Film. The first season
(12 shorts), launched in October 2006, captivated young
audiences who shared and published them through different
social media and the web; some people even created
their own shorts. Since then, Imaginantes has won numerous
awards, and it’s on the 4th season with a total of 46
shorts produced and delivered in several media (Imaginantes,
2006).
 1
An animatic is a visualization tool used in the pre-production
process of an animation that informs about movement, narrative
structure, framing aspects and visual effects to the production
team before the animation is actually done.
e-Motion is part of a research project on computational
creativity where we take a proven, creative process that
produces a recognized, valuable product and use it as an
environment to test our systems. We are interested
in studying how the overall creativity is affected if
computer systems take over different roles within the
pre-production stage. The Imaginantes team and process
are well defined as well as the work products that must be
produced. We hypothesize that all stages and work products
contribute to the overall creativity but we test our system’s
creativity by inserting it in the human process, to see
how it affects the outcome and ask human members of the
team to assess the system's performance.
There are two main advantages in this approach:
1. The system is assessed within a recognized real-world
creative process, so we can avoid the toy-world
generalization problem.
2. Our system plays a role within a human process, so it
is easier for the human members of the team to assess
the system's performance. They are experts in the
area, they know very well what to expect.
The following are the main motivations for our work:
• Study computational creativity within real-world
creative practices
• Understand the creative process of multi-sensorial
content, sound and movement of visual narratives.
• Develop a computational system that works
collaboratively in the creative production process of
visual narratives.
• Develop sound user criteria to evaluate the system as
a valuable tool.
• Experiment with automatically-created motion graphics
to study how different frame, color and graphic
element combinations transmit emotional content to
an audience. We want to maximize narrative appeal
while preserving the logical structure suggested by
the original plot. (Malamed, 2009).
Animatics constitute an intermediate step towards a full
motion graphics piece and, although they depict simple
Proceedings of the Fourth International Conference on Computational Creativity 2013 184
representations, they have a complex structure and most of
the high level architecture and elements of the final
product. They are built from storyboards,2 which in turn,
are assembled from scripts and constitute an important tool
within the process to develop motion graphics. They
convey decisions about editing, camera framing and
special effects. A production team can discuss several of
these options using various animatics before embarking on
the production stage, saving resources on this costly
process. Hart actually describes animatics as the “future of
motion control” to stress their importance (Hart, J. 2008).
As a starting point of our research into the nature of the
creation of animated stories, we use the output of Mexica
(Pérez y Pérez and Sharples, 2001), a computer system that
generates story-plots about characters, places and themes
of pre-Hispanic folklore; in particular, that of the Mexicas
(most commonly known as Aztecs). These stories were
originally represented in codices: pictographic documents
where cultures from Mesoamerica used to write their history
and other important aspects of their lives (Galarza,
J.1997).
Mexica plots are useful for our purpose because they
have very well defined and simple syntactic and narrative
structures, yet they have an immense potential for expression.
In fact most of the themes of classical literature can
be represented by Mexica plots: betrayal, sacrifice, courtly
love, deceit, loyalty conflict, etc.
The basic visual elements to assemble the animatic are
provided by another system: Visual Narrator (VN) (Pérez y
Pérez at al. 2012). This program illustrates story-plots from
Mexica by producing sequences of still images composed
of characters and scenes that literally represent the input
plot by following a set of rules used in some pre-hispanic
codices in a pictographic fashion. The rules specify how
characters are presented according to their rank in society,
activity, gender, and tension (emotional links represented
by facial expression). They also tell how locations must be
represented as well as action conventions. For instance, the
rules describe how to represent a person that has a high
social rank, who is talking to the people and who is angry.
All characters used by e-Motion are built by VN, within
the context of the process to produce a full motion graphics
piece, the sequence produced by VN can be considered as
a rough storyboard.
 e-Motion generates animatics that follow a set of conventions
for the representation of characters and locations,
but also depict the dynamics of the action and emotion
found in the original plot.
 2 Sequential drawings adapted from the script, depicted as concept
drawings that illuminate and augment the script narrative.
(Hart, J. 2008)
This paper is divided into four sections besides this introduction.
In the first section we describe the Imaginantes
project and why we think the use of computer systems can
improve it. In the next section we explain how the system
works. Then we propose a set of criteria to evaluate the
system. Finally we present some conclusions and the
current state of our project.
Building Motion Graphics for Imaginantes
Motion graphics are already present in everyday life. They
are used in a variety of media: TV identities, film titles and
credits, DVD’s, videogames, smartphones interfaces,
advertising displays and multiple media.
The creation of motion graphics is considered a special
skill, usually handled by artists or graphic designers
focused on the combination of design and television
broadcast or film (Frantz, M. 2003). The term is an
abbreviation of “Motion Graphic Design”. Kook refers to it
as the use of graphics, video footage and animation
technology to create the illusion of motion or rotation,
usually combined with audio (Kook, E. 2011).
The Imaginante’s team consists of 8 to 10 people
including: an executive producer, an art director, a design
and animation coordinator, animators, illustrators, a
musician or audio designer. The total time spent on the
creation of a short ranges from 10 to 12 weeks.
The team starts with an original script that provides the
general structure of the story. In some cases, this script has
some extra indications describing shots, special effects,
sound, etc.
Concept creation. The team collects all kinds of reference
material related to the theme and author. It’s a collaborative
and exploratory work.
Pre-visualization. At this stage, the team develops two
main tools: first, the storyborad, whose purpose is to show
the key moments of the story in a sequence, suggest
framing of the scenes and inform other specifics, like
lighting, camera movements and special effects. It gives
the entire pre-production team, a visual sequential
breakdown of the main scenes in the narrative.
The other tool is the animatic, which brings the storyboard
alive with motion, visual effects and a visual style for the
animation. It is very effective tool to pace the narrative and
timing and later add music and dialog (Hart, J. 2008).
Production. After the animatic is developed, illustrations
are created, digitalized, and rendered; sound and music are
also added to produce the final piece.
In a process like the one described above, a system like eMotion,
that suggests a variety of animatics with some
camera-direction decisions based on the dramatic content
the director wants to pursue, would be of great value for
the production team. In the regular process there is a limProceedings
of the Fourth International Conference on Computational Creativity 2013 185
ited feedback the team receives from just one animatic per
motion graphics project. It would also open new communication
channels between the team members by expanding
the discussion to new options and save time and work
resources.
e-Motion
Plots, in Mexica, are built by selecting characters and
structure from a repository of previous plots, combining
them in a way that makes sense, story-wise, and trying to
preserve well-known, successful emotional tensions. Emotional
tensions are collected during the process in an emotional-tension
profile for the story. This can be viewed as a
chart where overall emotion varies against time. Emotional
tension preservation in Mexica is a key factor in the
guidance towards the selection and combination of elements
for a successful plot.
A plot is a sequence of events in the order they occur in the
story. It is the skeleton, the structure that tells the main
events that occur in the story in a sequence of short action
descriptions. Before the story is complete and ready for a
final reader, it would have to be further developed to include
all aspects that fulfill a creative piece of literary
work. Yet, for our purposes it constitutes a good starting
point because in the Imaginantes process the starting point
is a plot from a text script (similar to a plot) with a few
very structured actions or events in sequential order.
An example story plot can be seen below. Emotional
tensions are inserted between brackets as they occur (Lc =
love conflict); (Lr = life at risk), (Hr = health at risk),
(Ad = actor dead):
Jaguar Knight was in Texcoco Lake
Enemy was in Texcoco Lake
Enemy got intensely jealous of Jaguar Knight (Lc)
Enemy Attacked Jaguar Knight (Lr)
Jaguar Knight fought Enemy
Enemy wounded Jaguar Knight (Hr)
Enemy ran away
Enemy went back to Texcoco Lake
Enemy did not cure Jaguar Knight (Lr)
Farmer prepared to sacrifice enemy
Enemy ran away
Jaguar Knight died by injuries (Ad)
Figure 1. A plot from Mexica and its tensions.
In e-Motion, a story plot with its emotional profile is taken
as input as well as a set of characters generated by VN. An
example character from VN can be seen in (Figure 2). It
depicts the 'enemy' character from the story being angry as
an emotional response to the fight it held with jaguar
knight (see plot above in Figure 1).
Each line in the plot is an event and these, in turn, are
incorporated into scenes by e-Motion. A scene has a set of
performers. A performer is a character in action. That is, a
character associated to an action to be performed. Characters
in the animation include anything that appears on the
screen and can be animated: humans, locations, emotional
tokens, etc. They are all images and can be modified by
'moods'.

Figure 2. Enemy in angry mood
A character can have several moods depending on the representational
variations available to it. A human can be
looking right or left (there are only two dimensions, so
far); he/she can be normal, angry or sad, etc. A location
can have rain, sunshine, etc. Actions encode the movements
of the characters on screen, they have a name and
are a combination of the following basic animation operations:
translation, scale and rotation. Performers can realize
an action from the plot, like 'fight' or enact the manifestation
of an emotional-tension like 'got intensely jealous of'.
Emotions in animation may be expressed in different
ways: character moods, textures flying as clouds across the
scene, icons depicting specific feelings —similar to the
ones presented in codices—, scene elements or characters
appearing as text, etc. In the latter case, the font, size and
color of the text are used to manifest different emotions
too. We call these: emotional tokens.
e-Motion builds scenes by following cinematic rules of
composition: transition, character distribution, motion trajectories,
framing and color. There are several options for
each and the system builds the scenes of the animatic by
selecting combinations of them that reflect the emotional
profile of the original plot.
Emotional tensions may be of different kinds: love, hate,
danger, anger, etc. As a story progresses, each event may
bring new emotional tensions into consideration. Some of
them may reinforce others previously introduced or may
counteract them. Each new tension introduced in a plot
manifest itself in the composition of a scene by affecting,
in a certain amount, the dramatic quality of the scene. For
every character there is an emotional profile consisting of
three parameters (Table 1): affect (the level of acceptance
/rejection the character feels), health (a level of well-being)
and excitement (a measurement of the degree of arousal of
the character). Each occurrence of a tension in the original
plot contributes, by a certain, predefined, amount to some
of the emotional profile parameters just mentioned. They
are ranked from -3 to 3. Hence, whenever a character is
to be integrated into a scene, its emotional profile defines
Proceedings of the Fourth International Conference on Computational Creativity 2013 186
how it appears in it: a character's affect, health and excitement
values determine its performance parameters in the
scene: mood, speed, and emotional tokens (Table 1).
There is also a set of global rules that determine framing
transition and trajectory for the characters and scenes.
Framing refers to decisions based on how close to
frame an action of the story and how far to pull it back so
the audience can see where the action is taking place. eMotion
choses its framing from a range of 4 types, based
on camera angles of photography and film (McCloud, S.
2006): First plane, Middle-shot, Middle close-up and Extreme
close-up. A story will always start with a first plane
view (a), making a zoom in camera movement, followed by
a middle-shot (b). As the story continues with the events
carrying on, the characters change their affect and health
levels. e-Motion will always look for the highest tension
value to change the framing. E.g. (+2 or -2) or (+3 or -3)
will change to frame (c) or (d) respectively. Every time we
have a middle close-up the program sets back to a middle
shot view until the levels of tension of the character are
increased again to +3 or -3 values. The system will always
end the story with a first plane view unless the overall level
of excitement shows -3 or +3. In that case the system selects
and extreme close-up and tilts the object in the frame.
Trajectory refers to the path that a character follows to
enter and exit a scene. Each character has an entrance to
the scene and a position to move to in the plane. It can also
continue it towards the edges of the plane.
Emotional tokens have two main trajectory paths: clouds
follow a curve; stains and pictograms stay in their initial
position of appearance while varying their scale according
to the character's level of excitement.
Transition refers to the sequence of movements from
one key scene of the story to another, thus establishing the
flow of the story (McCloud, S. 1993).
The example plot shown in Figure 1 produces the animatic
presented in Figure 3 as a sequence of images that
show some of its scenes. The third frame shows character
Enemy in angry mood because he is jealous of Jaguar
Knight. There is a green cloud traveling across the frame
from left to right showing that feeling in the scene. The
eagle and the cactus compose the name of “Texcoco
Lake”, the place where the scene takes place. The angry
mood is due to a low level of affect (-1) and excitement (-
1). The cloud is one of many possible manifestations of
jealousy; this one in particular was selected randomly. The
framing of that scene is solved in a middle shot angle. In
the last scene Jaguar Knight is dead. The tension refers to
Actor death, health (-3), excitement (0) and affect (0).
There is an emotional token (red stain) which depicts the
“being hurt”action and an up-scaling black cloud covers
the body, dissolving into the final scene.
Assesing Animatics
To evaluate an animatic we have designed a questionnaire
to measure its efficiency as a valuable tool and how new
and surprising are the results to the team. (Boden, M.
1992). The questions were designed following interviews
with actual members of the team where they set the parameters
for effectiveness. The questionnaire is aimed at
members of the Imaginantes team and will be rated higher
according to their experience. Each questions offers
5 levels of agreement (1 means “totally disagree”, while 5
Figure 3. Six still frames extracted from an animatic as a means of illustration.
Figure 3. Six still frames extracted from an animatic as a means of illustration.
Dimension -3 -2 -1 0 1 2 3
affection hate contempt envy neutral sympathy affection love
health death illness hurt neutral sane welfare happiness
excitement horror dread cautious neutral surprise joy bliss
Table 1. Dimensions of emotional profile and their discrete values.
Proceedings of the Fourth International Conference on Computational Creativity 2013 187
means “totally agree”):
1. Does the animatic show a logical selection of key
moments from the text script?
2. Does the animatic give you general information on how
many characters/objects/locations need to be drawn.
Does it help you visualize a particular graphic style?
3. Does the animatic give you a general direction on time,
special effects, movements and transitions you need to
consider for the animation?
4. Does the animatic show an appropriate selection of
camera angles according to the dramatic content of the
script?
5. Is it likely that anyone on your team would have come
up with a similar solution?
We end the questionnaire with a question that asks the participant
to locate in a chart the balance of intensity/clarity
and creativity of the animatic they have just seen
(McCloud, S. 2006). We expect that e-Motion should ideally
be ranked within the upper-right quadrant. (Figure 4).
All team members have access to all work products, so
they can relate them to the initial plot.
Clarity
Transmit the logical
structure narrative
of the text script
Intensity
Effects and techniques
to attract an excite
readers according to
the dramatic content
Very Creative
Suggests relevant ideas, some that
I have never thought before
Not so creative
Suggests a number of ideas but not so good
Figure 4. Efficiency and creativity chart
Conclusions and Future Work
e-Motion is a system that contributes to the development of
motion graphics by creating animatics. By doing so, it
plays an important role in the creative process since it determines
a great deal of the structure and action of the final
motion graphics piece. We consider our system important
because it allows us to experiment with computational creativity
in a proven creative process in the real world. This
setting is particularly appropriate, we find, to evaluate the
system's performance since the human part of the team can
do so, with well defined parameters. As far as the authors
of this paper are aware, there is no other work involving
computational creativity and animation.
At the time of writing e-Motion is in β -test for its first
version. We have run it with a few plots but the evaluation
process, although it has already been designed, it still
hasn't been applied. It will as soon as the system is ready.
In the first stage of the project we use Mexica plots as a
starting point. This allows us to use a standard for plots
that also contain information about their emotional tensions.
In the Imaginantes project, the starting point is a
script derived from an art piece. In subsequent versions we
will use the experience with Mexica plots to standardize
scripts taken from other sources and provide them with
emotional descriptions. We are also planning to develop
other systems that take over other aspects of the process
and study their effects.
The system currently works with a set of tension values;
these contribute cumulatively to character's emotional profiles,
which determine how characters are animated. The
general rules about framing, transition and trajectory
follow basic cinematic standards but we would also like
them to be determined by emotional parameters, an aspect
that needs to be further investigated
<references_biblio/>
References
Frantz, M. 2003. Changing Over Time: The Future of Motion
Graphics. Thesis and Research work. In
http://www.mattfrantz.com/thesisandresearch/motiongraph
ics.html.
Galarza, J. 1997. Los codices mexicanos. Arqueología
Mexicana. 4:6-24.
Imaginates 2006. In
http://www.fundaciontelevisa.org/imaginantes-2/page/3/.
Malamed, C. 2009. Visual Language for Designers: Principles
for creating graphics that People Understand. Massachusetts:
Rockport Publishers. 2252-3018.
Maher, M.L. 2012. Computational and Collective Creativity:
Who’s Being Creative? In Proceedings of International
Conference of Computational Creativity 2012.
Pérez y Pérez, R.; Sharples, M. 2001. MEXICA: a computer
model of a cognitive account of creative writing. Journal
of Experimental and Theoretical Artificial Intelligence.
Volume 13, number 2, pp. 119-139.
Pérez y Pérez, R.; Morales, N. and Rodríguez L. 2012. In
Proceedings of International Conference of Computational
Creativity 2012.
Hart, J. 2008. The Art of the Storyboard. A Filmmaker’s
introduction. Oxford, U.K: Focal Press. Elsevier 10:175.
McCloud, S. 1993. Understanding Comics The Invisible
Art. N.Y.: Harper Collins Publishers. 70-89
McCloud S; 2006. Making Comics: Storytelling Secrets of
Comics, Manga and Graphic Novels. Harper Collins Publishers.
19-25, 336-2900.
Proceedings of the Fourth International Conference on Computational Creativity 2013 188