ArticlePDF Available

Plausibility of 3D Characters: Towards a 2nd Uncanny Valley


Abstract and Figures

This paper overviews existing plausibility measures of 3D narrative animations with a focus on analyzing the realism of 3D characters and their performances. The main intent of this survey is to critically review factors that affect an overall perceptual sense of realism. Media literature addressing formal realism is connected to the technical facets of the 3D practice. This theoretical-technical bridge creates a case study database, from which a novel graphical system called The Character Plausibility Graph is proposed. The Character Plausibility Graph acts as an expanded edition of the well-known Uncanny Valley diagram, and depicts the relative effects of different formal visual elements on the character' sense of plausibility. The proposed graph is a dissection of a 3D character into its static and kinetic components and hence provides a visual tool to reflect upon the effects individually. This research concludes that stylization better serves the overall sense of plausibility than mere photo-realism or naturalism. It also reveals that performance of the characters and their animating method play a more significant role in the constitution of their realism, than their static design and rendering style.
Content may be subject to copyright.
Plausibility of 3D Characters: Towards a 2nd Uncanny Valley
Seyed M. Tabatabaei
Concordia University, Department of Design and Computation Arts, Montreal, Canada
This paper overviews existing plausibility measures of 3D narrative
animations with a focus on analyzing the realism of 3D characters
and their performances. The main intent of this survey is to
critically review factors that affect an overall perceptual sense of
realism. Media literature addressing formal realism is connected to
the technical facets of the 3D practice. This theoretical-technical
bridge creates a case study database, from which a novel graphical
system called The Character Plausibility Graph is proposed. The
Character Plausibility Graph acts as an expanded edition of the
well-known Uncanny Valley diagram, and depicts the relative
effects of different formal visual elements on the character’ sense
of plausibility. The proposed graph is a dissection of a 3D character
into its static and kinetic components and hence provides a visual
tool to reflect upon the effects individually. This research concludes
that stylization better serves the overall sense of plausibility than
mere photo-realism or naturalism. It also reveals that performance
of the characters and their animating method play a more
significant role in the constitution of their realism, than their static
design and rendering style.
Keywords: plausibility, realism, character performance, 3D
animation, uncanny valley
1 Introduction
In the field of 3D Computer-Generated Imagery (CGI or CG), it
can be confusing to understand what is meant by referring to a work
as realistic” when popular media examples of realism can be
wildly subjective in their interpretation.
This in turn makes it
difficult to know where in a CG narrative one should place their
efforts to maximize the sense of realism. Further, and the question
this paper seeks to unpack is : how important is the role of a
plausible 3D character in the construction of a CG narrative?
Throughout this paper, the term “plausibility” is mostly favored
over the more commonly debated term “realism”. Realism is often
muddled with the notion of “photo-realism” especially in the
context of CGI. Lev Manovich in The Language of New Media
(2001) argues that “what computer graphics has (almost) achieved
is not realism, but only photorealism the ability to fake not our
perceptual and bodily experience of reality but only its
photographic image.” [1] As an alternative notion, plausibility can
be defined as the viewer’s construction of a possible world whose
static and kinetic perceptual cues are in coherence and harmony,
and are intelligible to the viewers of different cultures with various
literacy of the cinematic language [2]. Therefore, Plausibility
accounts for the perception of the viewer and their level of the
engagement in the creation of realism effect. By definition, it
Why a CG character like Gollum in The Lord of the Rings (2001-203) is still
perceived a one the most believable characters, for example?
This was the central research question of my MFA thesis in animation, which was
conducted as a research through animation-creation project; Light Sight (2014). This
CG narrative animation centers around a single character and hence it brought the
entails other concepts like suspension of disbelief, immersion and
transportation which are relevant though not directly addressed by
this paper. David Surman in his paper CGI Animation:
Pseudorealism, Perception and Possible Worlds, relates
plausibility to the notion of perceptual realism from Stephen Prince
who suggests it as an alternative model of realism that relies on the
social correspondence and interpretations of the viewer [2]. As
such, “unreal images may be referentially fictional but perceptually
realistic”. [2-3] He expands this idea by noting that such perceptual
cues are not necessarily limited to visual (i.e. depth cues), but could
also be rooted in the social experiences of the viewer in reality.
In her Ph.D dissertation, Fatemeh Hosseini-Shakib [4] splits the
formal realism of the 3D medium into three categories:
Spatial(ontological), Cinematic(technical) and Character realism.
Her focus is on the realism of the 3D classic animations (i.e.
puppet-clay), which has many similarities with 3D digital
animation. This model allows the isolation of the character in the
study of realism and plausibility, which is a prime objective of this
research. The question is what aspect of 3D character design,
rendering and performance lead to the reality effect and potentially
leverage the viewer’s sense of plausibility.
But before focusing on the 3D characters, I explore first two other
categories that take part in the formal realism of any 3D animation
in general: spatial and cinematic realisms. Facets of these two types
of realism are intertwined with the realism of the 3D character. As
will be discussed, although all these types of realism could
contribute to the sense of plausibility, they do not necessarily
secure it or by no means are equivalent notions.
2 Spatial (Ontological) Realism
Shakib explains the ontological similarities between realism of
cinema and clay-puppet animations in three categories informed by
the work of well-known film and stop-motion theorists, Andre
Bazin and Michael Frierson. They are: Three-dimensionality
“meaning that the real and three-dimensional space is shown as it
is, using the same process as live-action cinematography”,
Objectivity “meaning that the images are photographs of tangible
objects in the real world” and Rawness “meaning that the medium
is transparent much as in live-action”. [3] Three-dimensionality is
an inborn quality of the CG since the very first feature that a 3D
software simulates is a 3D Cartesian space. Two other qualities also
exist in CG, however, the camera which used to be known as a
transparent recording device in capturing reality as it is, now is
replaced by the computer which acts as “kind of film camera and
film projector at once” [1]. This reminds us with two filmic
qualities that Siegfried Kracauer identifies as recording and
question of character realism to focus. However, this paper does not focus on the
analyses of this film.
Seyed M. Tabatabaei
revealing in cinema [4], which are now enacted by the process of
simulation and rendering in CGI.
2.1 Rendering
is the simulation of light behavior before and after
hitting a virtual object. All three mentioned ontological qualities of
a 3D CG world are constituted throughout the rendering process;
rendering informs the objectivity and tactility of the 3D entities by
processing the materials. The inherent three-dimensionality
coupled with elaborate lighting, shading and texturing make the 3D
characters potentially combinable with the live-action scenes; no
matter if the character design be abstract like Lego toys in The Lego
movie (2014), or fantasy like Smaug in The Hobbit series (2012-
2014), or real like Benjamin Button in The Curious Case of
Benjamin Button (2008) (Image.1). The more meticulous and
the render is, the less likely the final image gets mediated
and tweaked by the artist, and hence the more directly it would
resemble the photographic apparatus.
Image.1. 1. The Lego movie (2014), 2. The Hobbit series (2012-
2014), 3. The Curious Case of Benjamin Button (2008).
Surman describes this synergy as an “incarnation of co-
authorship”, first at an invisible level of mathematical algorithm
(programmer), and second at a visible level of subject
choice/creation (artist). [2] Incoherent collaboration could lead to
ambiguous styles and eventually collapse of the plausibility (this
will be elaborated in the stylization section). Other than casting the
ontological(spatial) qualities, rendering takes also an important role
in creating the cinematic realism of the CGI.
3 Cinematic (Technical) Realism
Cinematic realism for a 3D animation is the imitation of the
aesthetic codes and conventions of the cinema [4]. It is also referred
to as the Second-order realism by the media theorist, Andrew
Darley [5]. He coined this term to describe the realism of the early
Pixar’s short animations, which is drawn upon the established
grammar of the previous media (live-action cinema as well as 2D
animated cartoons of the Disney at the time). Second-order realism
could be seen as the dominant aesthetic policy of the Pixar and
other mainstream companies so far. Many facets of this kind of
realism is crafted by adopting what is known as the lens culture; it
includes effects like lens-flare, camera shakes and motion-blur,
depth of fields (DOF) and etc. Such optical aesthetics and the whole
mechanism of the virtual camera are simulated again within the
rendering process [2].
for more details, The Art of Rendering (Mike Seymour, 2012)
According to the book Special effects: the history and technique,
Innate competency of CGI in better simulating such cinematic
effects was a main reason why the digital version of the dinosaurs
in Jurassic Park (1993) was preferred over the puppet prototypes;
stop-motion animation is by its nature unable to produce motion-
blurs, and this limitation was preventing a seamless unification of
the real and puppet characters [6]. However, the resulted jerky
movements granted a unique aesthetic to the stop-motion animation
which has been counter-simulated in many 3D CGI animations too;
Room on the Broom (2012) and Allumete VR (2016) are good
examples of borrowing the aesthetic codes of the puppet-animation,
which by definition gives them a second-order realism. This is
technically done by avoiding the smooth 24fps movements,
applying some clay type materials and also minimizing the DOF
which gives the impression of small-size objects. Conversely, 3D
animations like Polar Express (2004), although full of realistically
designed characters, have been criticized for breaking the cinematic
realism, since their overtly free and smooth camera movements act
against the cinematic norms and constraints.
Relying merely on both spatial and cinematic realism via detailed
rendering brings about a sense of photo-realism, but does not
guarantee the overall plausibility. Besides that, photo-realism is a
highly technology-centric approach, as opposed to the stylized-
4 Stylization and Plausibility
Stylization could be described as deliberately tweaking some
aspects of the spatial, cinematic and character realism, for the sake
of an effective expressive communication. Stylization strategies are
taken to let the viewer participate more actively in the process of
meaning making and hence they could be more effective than
methods of photo-realism in creating the viewer’s sense of
plausibility. An extreme example of stylization is South Park (TV
series, 1997) (Image.2), whose three-dimensionality is almost
entirely abandoned, to the extent where it becomes
indistinguishable from a 2D cut-out animation. Stylization
definitely liberates the formal elements from complying with
photo-realism, and converge them towards a deeper sense of
Image.2. 1. South Park (TV series, 1997), 2. Eran Hilleli’s
experiments (, 3. The Extreme Wolds (2010,
In practice, stylization is attainable via abstracting, exaggerating or
deemphasizing some of the formal features. Well-known Indie
animators like Eran Hilleli (Image.2) and David OReilly take
minimal and surreal methods respectively and employ aesthetics
beyond the visual norms of the 3D medium. OReilly nicely
Un-biased and Biased rendering in computer graphics refer to the technical
classification of rendering mechanism based on the accuracy of the light simulation
calculated by the render-engines. (refer to Wikipedia)
Seyed M. Tabatabaei
summarizes this practice by saying that “the [stylization] rules you
set are as important as what you avoid using ... no motion blur, no
focal blur, no ray tracing or complex shading, no mesh-smooth, no
handheld camera, no vignettes, no glow, no fades, no crossfades,
wipes or transitions. [16].
It is noteworthy to add that stylization yields plausibility only if it
becomes intelligible by the target viewer. For instance, Final
Fantasy; The Spirits Within (2001), as Surman debates, fails to
communicate to the American audience, as effectively as the
Japanese viewers, since this animation signals the syntaxes of a
very specific Japanese genre: science-fiction anime [2]. So it
becomes implausible in American animation context which is
preoccupied with the Disney style. That is why Surman asserts that
“plausible requires affirmation by the previous discourse.” [2].
That said, even mainstream studios have also started adopting
stylization schemes against 3D aesthetic norms; Paperman (2012)
is a sample of Disney’s effort to revitalize its classic 2D cell look.
Image.3. 1. Fantasy; The Spirits Within (2001), 2. Paperman
5 Character Realism
Character realism is independently steered and affected by various
stylization approaches. However, the dominant culture of the
mainstream industry is to let renderings (lighting/textures) of the
characters as well as their surrounding environment be as photo-
realistic as possible, even in the most fantastical contexts. It is their
design (modeling) and/or movements that usually gets stylized.
This paradoxical alliance is to ensure that the characters are felt
ontologically real and stylistically engaging. Shakib uses the term
caricatured to denote the exaggerated design and appearance of the
characters, and cartoony style to signify their exaggerated
Caricatured characters are strongly associated with naturalistic
movements and have been criticized more often for their
implausibility, regardless of their sophisticated renderings.
Exemplars of this trend are the characters like Captain Haddock in
The Adventures of TinTin (2011), or Hiccup in How to Train Your
Dragon series (2010-2014). From Darley’s point of view, the
realism of such a trend is heavily dependent on fascinating the
spectator by means of the spectacle and sensuous effects [7-8].
Naturalism used to describe one of the Disney’s inherited style of animation which
adhered to the natural laws of physics. Albeit, it was mixed with exaggerated
movements, depending on the character’s personal traits and the story context. [2-3]
Image.4. 1. Hiccup in How to Train Your Dragon series (2014), 2.
Captain Haddock in The Adventures of TinTin (2011)
However, Surman contends that caricature is still a strong
expressive device which amplifies some culturally specific
perceptual cues and hence produces an intelligible communication
with the audience. Arguably drawing upon slight caricatured
stylizations saved a lot of plausibility for Captain Haddock,
compared to Dr. Aki in Final Fantasy; The Spirits Within (2001),
who lacks any kind of exaggerations, and turned to be an example
of implausibility.
Surman also asserts that the “affective power of the kinetic” should
not be underestimated as it carries the emotional signifiers.
Animated caricature is the term he uses for the stylized movement
which seems a better label compared to cartoony which implies a
certain set of stylization known as stretchy and squashy movements.
Animated caricature
“transcends still caricature; the performative.
‘Animate’ dimension allows for a greater complexity of
exaggeration…” [2].
Exaggerated movements should be afforded by the 3D model of the
character. This is delivered through the process of rigging.
5.1 Parameters of the kinetic stylizations
5.1.1 Rigging: affordance of style
Importance of rigging is usually overshadowed because of the
invisibility of its process and by the prominence of the animating
phase. It is the quintessential process which gives the 3d character
the flexibility to carry either a totally natural movement or to
burden highly stylized performances. Rigging has a medial status
as it partakes in both stages of character design and character
animation. In A Facial Rigging Survey (2012), rigging is
analogized with the control setup of the real marionette [17]; so
rigging design should afford the intended movement style and
restrict unnecessary modifications. It should ideally stay invisible,
but when it is poorly done or even so perfectly crafted but beyond
the narrative demands, it shows off. Thus, when the character starts
to be animated, bad rigging could break the anatomical integrity in
naturalistic approaches or visual appeal in stylistic ones,
The highest rigging challenge is the facial rig; it is usually designed
independently from other body parts due to its complexity. This
makes the Graphical User Interface (GUI) of this area typically
dense and unintuitive. Moreover, on top of the facial rig, are
sometimes added the Free Form Deformers (FFDs), for stylized
approaches or Physically-Based Deformers (muscles), for the
realistic demands [17]. Regardless of the complexity, the primary
kinetic stylization is used to prevent confusion and to give it a broader usage, within
which any sort of abstraction or exaggeration of dynamism is considered.
Seyed M. Tabatabaei
role of rigging design is to provide an interface that should
correspond to different motion data inputs by different agencies for
the sake of different movement styles.
5.1.2 Motion Input Data: multitude of agencies
Taking mainly from A Facial Rigging Survey (2012), coupled with
my own 3D practices, Figure.1 is arranged to visualize the agencies
that take part in movement creation in a CG environment. This also
creates a mind map for later discussions about character realism.
3D animation is perhaps the only form of media that invites a
multitude of agencies in generating the movements. This increases
the chance of style incoherence and misleading of the authorship.
It becomes especially evident when the character is fictitious in
nature yet acting in a realistic context (live-action), or should look
realistic but in a fantasy world. Hulk in The Incredible Hulk (2008),
and Knave of Hearts in Alice in Wonder Land (2010), are good
examples of both cases respectively; they both drew criticism of
plausibility in their own context. Generally, both Motion Capture
(Mo-Cap) and Procedural Systems data inputs are preferable for
simulations of natural/physical movements and visual effects.
However, stylized characters whose movement should also follow
aspects of naturalism call for collaborative methods of animation,
generated by the actor/actress, programmer (technical artist), and
animator. That said, the Key-frame data input remains the most
effective method when stylization matters.
5.1.3 Animator: prime author for kinetic stylization
As discussed, stylization which has a direct relationship with the
plausibility could be carved throughout the modeling, rigging and
animating progress. The animator is the prime mediator of the
kinetic stylization; as such, even the procedural systems which
were once embedded in the simulation phase and merely driven by
the algorithms, could now be creatively choreographed by the
animators. Rapunzel’s hair in Tangled (2010) is a good example of
a stylistically gestured and animated component of a CG character.
This flexibility became only possible via combination of both key-
frame animation and dynamic simulations. Such artistic controls
made the stylization of the character and its attachments coherent
Technically, kinetic stylization is conducted in either or both phases
of: temporal exaggerations and deformational exaggerations. The
first type is the common step that deviates the character’s
movement from naturalism towards expressionism, and the second
type intensifies the movements and/or gives characters a sense of
humor; excessive deformations in animations like Hotel
Transylvania (2012) look acceptable within their comedy genre
context. Again, Tangled (2010) which was also the first Disney’s
fairy tale 3D animation, is full of both types of exaggerations. Chris
Carter in the analyses of this film’s characters proves that kinetic
stylization is not only limited to the movements, and expressive and
story-telling poses could add to the character’s appeal while they
stay at their holds, but also complement their performance [9].
Image.5. 1. Tangled (2010), 2. Hotel Transylvania (2012)
Led by the animator, the main purpose of stylization is to constitute
an intelligible language to the audience, and consequently leverage
the overall plausibility. By definition, kinetic stylization is very
closely correlated with the notion of performance.
5.2 Character Performance and Plausibility
Significance of the performance as a perceptual cue in CG
animations is usually overshadowed by the intense tendency
towards photo-realism which centers around static perceptual cues
(light, texture, etc.). In Kinesic constructions: An aesthetic analysis
of movement and performance in 3D animation, Adam de Beer
posits another problem; the movement within the motion-picture
context, intersects and slips into other concepts like performance
and thus making it hard to analyze them individually [10].
According to de Beer, performance is a semiotic and meaning-
making device. He adds that performance is not limited to the
visible mise en scène entities including the characters; since the
camera movements and the movements derived for the editing
process could deliberately contribute to interpretation of the
moving-image, they also fall into the definition of performance.
However, in this research, performance of the 3D characters within
a linear narrative structure is under attention.
Film theorists David Bordwell and Kristin Thomson define the
narrative as a chain of events in cause-effect relationship
occurring in time and space. [11] They assert that character is the
main agency in shaping these causal chains. Ed Hooks in his book
Acting for Animators (2013), accounts the motivated movements or
gestures (calling them the actions) as the essential components of
the character performance [12]. As such, movement on its own does
not necessarily lead to the performance, also its absence does not
equal the lack of performance. Hooks evokes scenes from Toy Story
film series where the characters do not move a single bit, yet as
their poses are interpreted motivational, they are considered
actions. Likewise, movements derived from the natural forces do
not contribute to the performance, no matter how much appealing
they appear, unless intentionally choreographed.
As opposed to the real characters in a theatrical/live-action context,
animated character’s actions should also foster the illusion of life.
In Hooks’s view, this is essential yet again not sufficient to make a
Seyed M. Tabatabaei
character plausible. He cites Hans in Frozen (2013) as a blank slate
that feels alive and real, but failing at serving the narrative’s causal
chain, thus hindering any empathy with him. In this way, Hans
becomes implausible at a much more profound level [14]. The
illusion of life and feel of empathy are two sides of the same coin;
a character’s plausibility. For CG characters, such deficiencies
usually lead to a harsher consequence, plunging them into the
Uncanny Valley.
Image.6. 1. Woody in Toy Story (1999), 2. Hans in Frozen (2013)
5.2.1 Uncanny Valley: question of humanlike characters
The concept of Uncanny Valley was first coined by Mashahiro
Mori, the Japanese roboticist [14]. It suggests that the more realistic
the humanoid robot looks, the more empathy it gets from the
perceiver, but up to a certain point beyond which it starts to feel
creepy and unfamiliar. As the graph shows, the added movement
increases the affinity but also makes the fall more severe. This
graph commonly used to explain the realism of CG characters as
well; if the movements do not breathe life to the characters, they
will convert them to zombies. Dr. Aki in Final Fantasy, has been a
classic example of such failure. Evidently, Dr. Aki and almost all
of the 3D characters who confronted this problem, relied on the
Mo-Cap techniques.
We can also discern that our level of expectation impacts the way
we perceive the realism of a 3D character. There are factors altering
our expectations; for instance, a nuance mistake in performance of
a human-like character in a real context could undermine the whole
plausibility, whereas the same type of error for an abstract character
in a fantasy world entails less challengeseven though both having
a same level of rendering and movement details. We can suggest
Figure.3 to illustrate levels of expectation for various genres of
characters. Gollum (6) for instance, provokes less reality-
expectation than Benjamin Button, since we tend to perceive him a
creature (intentionally uncanny) rather than a human who lives in a
fantasy world, unlike Benjamin Button who is peculiar yet human
and supposedly lives in our human world.
There are also other factors that affect the audience’ expectations;
De Beer, divide character’s body movements into two groups:
robust (body limbs) and fine (facial and fingers) movements, and
posits that fine movements have higher impact on the realism. [10].
Because they are the main communication tools, we less tolerate
their flaws. He also correctly relates the shot-framing to the way we
perceive robust and fine movements; i.e. close-up shot from the
face makes the eyes’ saccadic movements more salient and
noticeable. Importance of the facial performance on the plausibility
is well demonstrated by Andrew Buchanan [15]; he analyses the
facial expressions of the animated characters in three groups;
deliberate (with visible motivations), spontaneous (unconscious)
and symbolic (stylized and communicable to wider cultural
groups). He proves that spontaneous expressions are the most
effective in provoking the emotional and empathic responses
particularly in a dramatic narrative.
Based on what we discussed about the stylization approaches to 3D
realism, role of character’s performance, notion of uncanny valley
and audience’ plausibility expectations, we will now propose the
Plausibility Graph for 3D characters, which reveals another
uncanny-implausible valley.
1.BB-8/ Star Wars: The Force Awakens (2015)
2.Driller/ Transformers: Dark of the Moon (2011)
3.Brains/ Transformers: Dark of the Moon (2011)
4.Sonny/ I, Robot (2004)
5.Caesar/ Dawn of the Planet of the Apes (2014)
6.Gollum/ The Lord of the Rings Trilogy (2001-2003)
7.Gwaihir/ The Lord of the Rings Trilogy (2001-2003)
8.Richard Parker/ Life of Pi (2012)
9.T-800/ Terminator Genisys (2015)
10.Benjamin Button/ The Curious Case of Benjamin Button (2008)
Fig.2 Uncanny Valley
Seyed M. Tabatabaei
6 Plausibility Graph for 3D characters
Fundamental visual components of the CG character are framed on
the Figure.4. Performance, Design and Rendering (P/D/R) are the
visual elements whose accumulation shape the overall aesthetic of
the character and more importantly its realism/plausibility. As
discussed, the rigging plays an important part in defining the
performance style. Besides, dynamics of the character’s belongings
and accessories (i.e. hair, cloth, …) could deliberately contribute to
the meaning making (performance). However, P/D/R are the three
most explicit components that direct the visual style of the character
and each could separately lean towards realistic or stylized
approaches. Given that, a three-dimensional Cartesian graph is
imagined (Figure.5 TopView, Figure.6 IsometricView) whereby
the methods of movement are ordered on the X-axis and the static
design and rendering approaches are arranged on the Y-axis.
Accordingly, certain characters (mainly the protagonists) are
isolated and located relatively on the X and Y coordinates. The Z-
axis is assigned to the relative sense of character’s plausibility
(Fig.6). As argued, plausibility of the whole piece is a comparative
intricate concept that is affected by many other factors than solely
the character realism and hence it would be impossible to be
converted into a precise quantitate value. That said, values of the Z
axis that are represented as a fluid surface (Figure.6) are not
absolute. Characters were mostly chosen from the linear narrative
context; especially those which are positioned on the critical points
have been frequently debated by the critiques for their realism and
plausibility of the film they partake.
It is notable that to facilitate a reasonable comparison, only human
characters are selected; they dominate over the non-human
characters (animals, creatures and anthropomorphized objects)
both in quantity and style variations. Also as illustrated in Figure.3,
they are usually most at risk of implausibility.
6.1 Discussion (On X, Y Axes)
As depicted in Figure.5, movements of the characters on the right-
side are more naturalistic, and this gets even more evident when the
motion capture gradually replaces the key-framed method of
animation. On the left-side, performances are more stylized in
timing and stretchy/squashy exaggeration is added and hence the
cartoony effect is amplified. On the top quadrants, the renderings
and particularly the textures become more sophisticated and the
characters are potentially qualified to get composited into live-
action movies. Caricature approach is gradually omitted towards
the topmost area in which reside the hyper-real characters (who are
supposed to be identical to the real human, design and rendering-
wise). All the characters on the lower part have sorts of stylized
design towards abstract forms on the lower half, wherein the
stylized renderings also take part in creating either a clay-puppet
look or a 2D aesthetic.
3D Character
Non-Human (Animals-
Hybrid (realistic)
Motion Captured (MOCAP)
Secondary Motion
parts (Cloth,Hair,...)
Hybrid (Sylized)
Seyed M. Tabatabaei
Fig.5 Plausibility Graph: Arrangement of the characters
based on: Performance / Design / Rendering (P/D/R)
Seyed M. Tabatabaei
6.2 Discussion (On Z Axis)
As illustrated in the Figure.6 (isometric view of Fig.5), there are
two expected critical points that show in warmest colors; the deeper
decline which located in zone B evokes Mori’s Uncanny Valley
where the characters like Dr. Aki and Beowulf (B1,2)
are situated.
This graph, however, suggests that there seems to be another valley
traced in the zone D whose characters are also at risk of
implausibility, though they have a stylized look. Shared among
both areas is the utilized methods of movement which are less
mediated by the animator and rely more on the motion capture
DJ Walters (D6) and other characters in Monster House (2006) look
eerie, however their weird style deliberately complements the
mysterious mood of the whole film, and hence do not harm the
overall plausibility as much as Rostam(D7) because of its uncertain
style of rendering and unappealing design. Conversely, Deunan
Knute (D10) follows a more concrete visual style, imitating a
drawing expression and keeps distant from this 2nd uncanny valley,
though it was still animated by the Mo-Cap technique. Other shared
features between the two uncanny points is the badly constructed
facial expressions. As explained, although there is a high reality-
expectation around the fine performances (face and digits), there is
often a lack of coherence between facial and whole body animation
style in those two areas. As depicted, there are other potentially
critical areas roughly in the middle and top left, in which the
stylization direction is somehow uncertain and ambiguous.
Characters of such areas like Mowgli (D1) are perceived as
unappealing. For M.E.(D11) in Light Sight (2016), I tried to avoid
these areas by implementing a clear abstracted form inspired from
the design of puppets and androids.
Zone A, is where superheroes exist, due to the realistic
rendering/texturing and exaggerated movements. Characters of this
Coordinates of the mentioned characters are addressed as such; B1 means Character
Number. 1 on Zone B. Fig 5
area cannot be totally detached from the physical laws of
movements, and that is why characters like Hulk (A5) or Mr.
Fantastic(A1) are sometimes criticized for their excessive moves
and hence standing apart from their live-action context. For the
same reason, plausibility of the character performance in this zone
relies on a harmonious synergy of the key-frame animation and
motion-capture technique. It is noteworthy to mention that since
Hulk is read almost as a creature in its extreme states, it keeps a
safer distance to implausibility as opposed to Mr. Fantastic who is
always human(like) (refer to Figure.3 arguments).
This graph also tracks the overall scheme and style preferences of
the authors. Many of the Disney and Pixar characters stay on the
upper/right side of the zone C, whereas Sony Pictures gradually
diverted towards the left/lower side of the same zone. Directors like
Steven Spielberg, James Cameron, Peter Jackson and Robert
Zemeckis have perpetually preferred the characters of the zone B.
Recently, designing super-abstract characters with hyper-real
movements has become a popular trend and are well received by
the indie artists; capturing movements of the Dancing characters
(D13) in AICP Sponsor Reel (2016), is achieved from a
combination of advanced motion-capture technologies, procedural
animations and dynamic simulations.
This research overviewed the critical aspects of the realism in CG
animations in three venues of spatial, cinematic and character
realism. Although realism and plausibility are interdependent
concepts, plausibility was favored as it more explicitly relies on the
film’s credibility based on the perception of the viewer rather than
a precise indexical reference to the real (photo-realism). Proposed
Plausibility Graph well supports the idea that stylized-realism could
more controllably contribute to the sense of plausibility and at the
same time lead us to the novel ways of experiencing the realism.
This graph also proves that kinetic stylization informed by the
character’s performance, has a higher effect on the plausibility of
the characters rather than the static stylizations shaped by the
design and rendering approaches.
As depicted and discussed, technology plays an important role in
devising the production techniques, shaping the aesthetic outcomes
and affecting the realism of the CG characters. Ever increasing
advancements in performance-capture and real-time rendering
technologies, are gradually bridging the Uncanny valley(s). Both
animators and live-action actors/actresses have been often told that
their job is threatened by motion capture technologies. However,
these are in fact expanding the actor’s embodiment in various
avatars of different flexibilities beyond their own physical
limitations. Equally, such technologies are liberating the animators
form the realistic approaches dictated by the naturalistic school of
thought. This is somewhat how photographic camera once freed the
painters from hyper-real mimicry and consequently a diverse range
of stylized and abstracted approaches were flourished. And this
reminds us that no matter which gaps the technology fills,
implausibility valleys seem to always call for their own unique
Seyed M. Tabatabaei
I would like to thank Dr. Jonathan Lessard for supervising this
paper; Rebecca Goodine and Shaney Marie Herrmann for editing
and feedbacks; Dr. Fatemeh Hosseini Shakib for supervising the
original MFA thesis; faculty members of Tehran University of Art-
Department of Cinema & Theatre for supporting this project; Simin
Farrokh Ahmadi for assisting in making the graphs and arranging
the filmography.
Manovich, L. (2001). The language of new media. MIT press, 181, 47
Surman, D. (2004). CGI Animation: Pseudorealism, Perception and Possible
Worlds., 11, 38
Hosseini-Shakib, F. (2009). The hybrid nature of realism in the Aardman
studio’s early animated shorts (Doctoral dissertation, University of Brighton).
76, 80, 81, 101
DEMIDOV, V. E., HANSEN, U.-H., AND DEMOKRITOV, S. O. 2007. Spin wave
confinement. Phys. Rev. Lett. 98, 157203.
Darley, A. (1997). Second-order realism and post-modernist aesthetics in
computer animation. A Reader in Animation Studies, Sídney: John Libbey,
Rickitt, R. (2000). Special effects: the history and technique, 125
Darley, A. (2002). Visual digital culture: Surface play and spectacle in new
media genres. Routledge.
Kenny, C. (2014). How to Train Your Dragon 2: When Animated Films No
Longer Appear Animated. Retrieved April 08, 2018,
Carter, C (2013) An Analysis of the Character Animation in Disney’s Tangled,
Sense of Cinema [Online] Issue 67,
De Beer, A. (2010) Kinesic constructions: An aesthetic analysis of movement
and performance in 3D animation, Animation Studies Online Journal [Online]
Bordwell, D. & Thompson, K. (2008) Film Art: An Introduction (8th Edition)
Boston, [Mass.]; London: McGraw Hill.
Hooks, E. (2013). Acting for animators. Routledge.
Hooks, E. (2014) Disney’s ‘Frozen’: The Acting and Performance Analysis, [Online],
Mori, Masahiro, MacDorman, Karl F. and Kageki, Norri (2012) The Uncanny
Valley [from the field], Robotics & Automation Magazine, IEEE [Online]
vol.19 no.2 (p.98-100),
Buchanan, A. (2009). Facial expressions for empathic communication of
emotion in animated characters. Animation Studies, 4, 75.
OReily, D. (2009) Basic Animation Aesthetics, [Online],
Orvalho, V., Bastos, P., Parke, F. I., Oliveira, B., & Alvarez, X. (2012). A
Facial Rigging Survey. In Eurographics (STARs) (pp. 183-204).
Simmons, M., Ward, K., Yosumi, H., Leo, H., & Zhao, X. (2011, August).
Directing hair motion on Tangled. In ACM SIGGRAPH 2011 Talks (p. 41).
1. A Christmas Carol (2009) Robert Zemeckis, US
2. A Scanner Darkly (2006) Richard Linklater, US
3. AICP Sponsor Reel(2016) Rupert Burton, Method Studios, US
4. Allumette. Dir. Eugene Chung. Penrose Studio, 2016
5. Alice in Wonderland (2010) Tim Burton, US
6. Angelo Rules (TV Series 2010- ) Chloé Miller/ Franz Kirchner, France/
7. Appleseed Ex-Machina (2014) Shinji Aramaki, Japan
8. Asterix: The Land of the Gods (2014) Alexandre Astier/ Louis Clichy,
France/ Belgium
9. Avengers (2012) Joss Whedon, US
10. Beowulf (2007) Robert Zemeckis, US
11. Cloudy with a Chance of Meatballs (2009) Chriss Miller/ Phil Lord, US
12. Contre Temps (2012) Lucas Veber/ Thibaud Clergue, France
13. Dawn of the Planet of the Apes (2014) Matt Reeves, US
14. Despicable Me (2010) Pierre Coffin/ Chris Renaud, US
15. Fantastic Four (2005) Tim Story, US/ Germany
16. Final Fantasy: The Spirits Within (2001) Hironobu Sakaguchi, US/ Japan
17. Frozen (2013) Chris Buck/ Jeniffer Lee, US
18. Galaxy ‘Dove Chauffeur’ Advertisement (2013) Daniel Kleinman,
Rattling Stick, US
19. Hotel Transylvania (2012) Genndy Tartakovsky, US
20. How to Train Your Dragon 2 (2014) Dean DeBlois, US
21. Hulk (2003) Ang Lee, US
22. I, Robot (2004) Alex Proyas, US
23. Jurassic Park (1993) Steven Spielberg, US
24. King Kong (2005) Peter Jackson, New Zealand/ US
25. Life of Pi (2012) Ang Lee, US
26. Light Sight (2016) Seyed Moslem Tabatabei, Iran
27. Maleficent (2014) Robert Stromberg, US
28. Monster House (2006) Gil Kenan, US
29. Mr. Peabody & Sherman (2014) Rob Minkoff, US
30. Panda Bear “Boys Latin” (2014) Isaiah Saxon/ Sean Hellfritsch, US
31. Paperman (2012) John Kahrs, US
32. Passion Ski (2010) Jean-Nicholas Arnoux/ Quentin Baillieux, France
33. Please Say Something (2008) David O’Reilly, Germany
34. Popeye (2016) Genndy Tartakovsky, US
35. Pythagasaurus (2011) Peter Peake, UK
36. Rise of the Guardians (2012) Peter Ramsey, US
37. Room on the Broom (TV Short 2012) Max Lang/ Jan Lachauer, UK/
38. Shrek (2001) Andrew Adamson/ Vicky Jenson, US
39. Spider-Man 3 (2007) Sam Raimi, US
40. Star Wars: The Force Awakens (2015) J. J. Abrams, US
41. Tangled (2010) Nathan Greno/ Byron Howard, US
42. Terminator Genisys (2015) Alan Taylor, US
43. The Adventures of Tin Tin: The Secret of the Unicorn (2011) Steven
Spielberg, US/ New Zealand
44. The Backwater Gospel (2011) Bo Mathorne, Denmark
45. The Curious Case of Benjamin Button (2008) David Fincher, US
46. The Gruffalo (TV Short 2009) Max Lang/ Jakob Schuc, UK/ Germany
47. The Hobbit: An Unexpected Journey (2012) Peter Jackson, New Zealand/
48. The Incredibles (2004) Brad Bird, US
49. The Incredible Hulk (2008) Louis Leterrier, US
50. The Jungle Book (TV Series 2010- ) Tapaas Chakravarti, India
51. The Lego Movie (2014) Phil Lord/ Christopher Miller, US/ Australia/
52. The Lord of the Rings Trilogy (2001-2003) Peter Jackson, New Zealand/
53. The Mask (1994) Charles Russell, US
54. The Matrix Reloaded (2003) Andy Wachowski/ Lana Wachowski,
Australia/ US
55. The Peanuts Movie (2015) Steve Martino, US
56. The Polar Express (2004) Robert Zemeckis, US
57. The Smurfs 2 (2013) Raja Gosnell, US
58. Toy Story (1995) John Lasseter, US
59. Toy Story 2 (1999) John Lasseter, US
60. Toy Story 3 (2010) Lee Unkrich, US
61. Transformers: Dark of the Moon (2011) Michael Bay, US
62. Up (2009) Pete Docter, US
63. Van Helsing (2004) Stephen Sommers, US
64. WALL-E (2008) Andrew Stanton, US
... Para Hito Steyerl (2014) es, tanto gráfica como metafóricamente, la mirada hegemónica del hombre blanco que otea el horizonte colonial desde su barco esclavista. Y las comunidades virtuales 3D, que pudieron construirse sobre distribuciones multiperspectiva y visiones simultáneas y múltiples (Farocki, 2013), optaron, directamente, por simular el espacio cartesiano y su historia de dominio escópico (Tabatabaei, 2018). ...
Conference Paper
Full-text available
El diseño de entornos, objetos y cuerpos digitales en 3D crece como estética preferente de artistas transfeministas y queer en los últimos tiempos. Las causas son difíciles de definir con certeza, pero podemos especular dos hipótesis. La primera, la no-necesidad de un referente real permite al modelado 3D crear un mundo desde cero. Consecuentemente, obviar una realidad notablemente violenta para los cuerpos no-normativos, y poder imaginar otro entorno de emancipación y libertad, parece favorecer este camino creativo. En segundo lugar, cuando el sujeto de nuestra resistencia política deja de ser binario, exclusivamente humano, de identidad fija y estable (Haraway 1985), entonces nuestras políticas de la representación exigen urgentemente una nueva estética que nos de cabida. Un cuerpo que se moldea sin referente, que no está atado a las reglas de la física clásica, que puede modificarse en cualquier momento y al que se puede acudir desde perspectivas múltiples y diferentes a la cartesiana (Stone 1997); puede solucionar, al menos momentáneamente, esta urgencia estética. Mi objetivo en esta comunicación es compartir los puntos de partida teóricos de mi tesis doctoral en realización: Especulaciones digitales, ficciones volumétricas. La imagen digital volumétrica/3D en el debate transfeminista contemporáneo. Comenzaré, precisamente, por especular las posibilidades del modelaje 3D para la creación transfeminista y queer en los dos sentidos abiertos previamente. Para ello, manifestaré la necesidad de abordar las políticas de la volumetría (Rocha y Snelting 2017) desde perspectivas transfeministas. Empujando dichas políticas de la volumetría, continuaré por señalar los grandes regímenes normativos que perpetúan las desigualdades estructurales y sistemáticas en los entornos virtuales tridimensionales, tales como la tradición militar (Virilio 1986), el cartesianismo (Stone 1997), el hiperrealismo, los regímenes verticales (Steyerl 2011) o la desigualdad de acceso y alfabetización digital. Por último, el cuerpo principal de la comunicación se situará en la posibilidad de una reapropiación transfeminista y queer de las herramientas del modelaje digital contemporáneo. Acudiré a diverses artistas contemporáneos que utilizan el 3D como su herramienta estética –Jacolby Satterwhite, Lu Yang, Tabita Rezaire o Johanna Bruckner, entre otres– para ejemplificar la posibilidad especulativa volumétrica. Junto al trabajo teórico de José Esteban Muñoz en Cruising Utopia (2009), profundizaré en las prácticas de fabulación, construcción de futuros y ciencia ficciones de sus piezas artísticas, como “practicas representacionales que nos ayudan a ver lo todavía-no-consciente”, y como “semillas de potencialidad” hacia los futuros transfeministas y queer que construimos y debatimos en nuestros movimientos políticos, afectos, deseos y reflexiones.
Full-text available
The challenge of communicating emotional content to an audience via animated characters has existed since the art form first appeared. As animation techniques and technology have advanced, animators and character designers find themselves with a multitude of resources and tools for the creation of facial expressions so as to effectively communicate the emotions of their animated characters within each scene. However, by using evolved forms of symbolic facial expression, which are widely accepted, these technologies and techniques often overlook the unconscious communication conveyed via actual human facial expressions. The fundamentals of this instant and unconscious emotional communication have been well studied and documented, yet the systems developed by the scientific community for reading and interpreting facial communication have only occasionally and recently been applied by animators.
State-of-the-art simulation techniques can produce compelling and natural hair motion. In creating Disney's feature film Tangled, physically plausible motion was important but just one foundational component. The story contains an unprecedented amount of interaction of the hair with the characters, as well as a high level of art-direction. Through 2D "drawovers" the artistic vision was conveyed in an often very detailed way - how the hair should move, and what poses and silhouettes it should hit. Simulation alone is not sufficient when such a high degree of direction and interaction is required. We describe a hybrid approach leveraging the power of custom hair dynamics with the artistic control of key-framed animation that was key to the success of directing hair motion on Tangled.
The hybrid nature of realism in the Aardman studio's early animated shorts (Doctoral dissertation
  • F Hosseini-Shakib
Hosseini-Shakib, F. (2009). The hybrid nature of realism in the Aardman studio's early animated shorts (Doctoral dissertation, University of Brighton). 76, 80, 81, 101
Spin wave confinement
  • V E Demidov
  • U.-H Hansen
  • S O Demokritov
DEMIDOV, V. E., HANSEN, U.-H., AND DEMOKRITOV, S. O. 2007. Spin wave confinement. Phys. Rev. Lett. 98, 157-203.