this is the interblag of ole christian alfheim. he feeds upon the souls of people watching his animations.

Virtual Camera in Classic Animation

“The Old Mill” (Wilfred Jackson, 1937) is a milestone in camera simulation in 2D animation as it is the first to take the use of a multi-plane setup that allows refined camera movements. This setup consists of up to seven glass layers of artwork and a moveable camera, this setup made it possible for a more versatile use than the Fleischer’s setup, as it was able to simulate that the camera was moving into and through the painted background. As in the Popeye short, “The Old Mill” is also using backgrounds that are painted with warped perspective to simulate turning or tilting. But it does it less subtle and more as an active part of the narrative, as with the crane shots from “Love Me Tonight”.

Right from the opening scene they show us the both potentials and limitations of the multi-plane cameras. While the virtual camera dollies in to the picture through the landscape and over the water, a shot that would be hard to duplicate in traditional cinema at that time, it still has to jump cut forward in motion four times to come to the end of the camera motion in to the picture, this could be seen as an artistic choice, but it had to be made because of the physical limitations of the multi-plane camera setup (as the physical camera cannot move through the glass plates that makes up the layers). Probably in an effort to enhance the feeling of depth and 3D space, elements in the scene comes out of focus as it gets close to the camera. This helps the audience to except the reality they see on screen, as it reminds them of how reality is depicted through a conventional camera.

Most notable is the third scene in the movie, where the virtual camera moves around inside the old mill. At first it’s tilted down moving upwards, and then it tilts up and dollies forward through the beams, before moving and tilting upwards. Some of the animals in this scene are responding and somewhat aware of the virtual camera and reacts to its presence. An interesting discovery is that although the movements of virtual camera in theory could be more floating, they are not, but they rather mimicked the style of the rigid movements the camera in traditional cinema at the time (tilting, turning, dollying and paning as separate movements). A possibly to reasons for this is that most of these movements were illusions in 2D space, and by not separate the camera movements to something the audience is already familiar with from cinema, the camera movements could be interpreted as confusing or in worst case not interpreted at all.

Virtual Cinematography in 2D animation

“Popeye the Sailor meets Sinbad the Sailor” was one of the animated shorts that used the Fleischer’s patented “Rotograph” camera setup. It used a horizontal multiplane camera setup, where the background (as well as middleground and forground) for the cartoon is actually modeled 3D miniatures in the same scale as the drawn cartoons. This gives a real stereoscopic depth of field when panning, also known as “parallax”.

It is clear that the cinematic narrative as found “Love Me Tonight” is evident in “Popeye the Sailor meets Sindbad the Sailor”, but the animated short adds further dynamic to the camera movements as it is not restricted to constructions such as a crane or a dolly to move the viewpoint in space.

In 2D animation, camera tilts and turns are done by using warped perspective in the background paintings, so when they are moved under the physical camera, you get a simulated tilt or turn. Such as in the opening scene of this animated short, where we are first introduced to island of Sindbad.

In the second shot of the first scene, we are presented with a camera tilt up from the “welcome sign” on the island, but the virtual camera is also moving in to the picture as it is tilting. Both physically by moving the real camera closer to the cels, and virtually by panning over the warped perspective on the painted background. The camera is following the winding road up to the castle for the first ten seconds of the scene, before slowing down its movement in depth and only continues tilting up to show the castle on the mountain top.

We are then presented with a long shot of the castle, a similar shot as to the long shot of Nosferatu’s castle in “Nosferatu” (F.W. Murnau, 1922). Then a medium shot from the same angle followed up by a close shot where the perspective is slightly different, as the angle of the viewpoint has slightly changed. A minor detail that enhance the illusion of a three dimensional world. In the following scene the camera follows Sindbad (the antagonist) as he walks across his island. Since the backgrounds are miniature models, the camera movement gives us a stereo-optical effect with a deep sense of depth and parallax, further enhancing the illusion of a three dimensional “reality”.

In the scene where Sindbad discovers Olive Oyl through his scope, we see the iris focusing in on Olive. It has a clear resemblance to the iris focus technique used in one of the romantic scenes in “Birth of a Nation” (D.W. Griffith, 1915).

The third time we see the evil bird fly in the sky, 6 minutes and 30 seconds into the movie, we see an interesting use of simulated camera. It is a mid shot of Sindbad’s evil bird who is flying in a circle up in the air screeching at Popeye’s ship (that is out of sight in this shot), while the camera is stationary it tilts and turns following the flight pattern of the bird. But the camera does not do so perfectly, it lags behind as if it is a filming of a real bird filmed by a cameraman who is a bit uncertain about the pace and direction the bird is flying, emulating the style of a shot from a nature documentary. By simulating the weaknesses of a real camera man in the movements of the virtual camera, it adds a layer of reality to the shot making it more believable as it fit with conventions the audience already know from the “reality” of nature documentaries in cinema.

These examples shows that the directors of traditional animation where mimicking the cinematography of different genres from cinema. By using a cross pollination of different from different genres in a production, they can play on conventions and grammar that the audience already know, and recall their associations to them in different shots as they see fitting by the narrative of the animation. By playing on conventions that are already being accepted as cinematic reality, they increase the believability of the constructed world.

Since the basic nature of the 2D animation in lacks real depth, I discovered that the animators compensate by using different techniques to give the constructed world a sense of three-dimensionality. Examples of these techniques are changing the angle of the viewpoint between shots, creating a parallaxing effect when panning, let action take place along the Z-axis of the viewpoint and building up a scene by positioning the action in the middleground while constructing the scene around it (in the foreground and background).

Liberated Camera View in Classic Cinema

Being made the same year as the invention of the zoom lens, “Love Me Tonight” contains the first zoom shot in a feature film. In the opening scene, when establishing the early in the morning in Paris, the camera zooms in on a couple of chimneys. At the end of the zoom, the chimneys let out smoke, before the movie proceeds to the next shot of the scene.

“Love Me Tonight” is using the crane to liberate the camera into movements in space. In the scene where we are introduced to the protagonist, the camera floats mid-air through his window into his bedroom. It is as possesses a life on its own, it is an active part of the narrative. This effect is used through the movie as the camera rather tilts and turns from a situation to the next. Most notable is this effect in the scene at the chateau, where the camera floats through the bedroom and then tilts down the sleeping protagonist, as through the eyes of a person moving through the room.

What these examples shows is that in classic cinema the films took advantage of new technical solutions as they got available to them, and used them to explore the camera narrative as well as to liberate the camera in terms of spatial movements. This shows that the idea of a free roaming camera existed in the 1930s’ cinema.

“Love Me Tonight” also employs effects such as slow- and fast-motion shots. Fast-motion is used to increase the feeling of intensity of when the princess is riding her horse to catch the train. Slow-motion is used in the scene when all the horse riders are on their way to ride offscreen, to enhance the dramatic shot of all the galloping horses and to give the viewer time to read what is happening onscreen.

This shows that it was a clear distinction between a special effect shot and regular shot also in classic cinema.

Historical Antecedents of the Animated Camera

“Gertie the Dinosaur” follows the style of the early cinema (also known as primitive cinema), which had a static camera where all the actions is played out in a very frontal style (“Frankenstein”, J. Searle Dawley, 1909), not unlike traditional theatre. An interesting point is that since McCay didn’t use cels in this short, and therefore had to redraw everything on screen on each frame, he had no restraints keeping him from creating a simulated camera movement, but he kept to the presentational style of primitive cinema. It is also interesting that the world he constructed is three dimensional many of the actions taking place along the Z-axis (in the depth of field). As with silent movies of that time, McCay uses black slides with white text as intertitles, to do the dialogue.

While “Gertie the Dinosaur” follows the style of early cinema, “The Sinking of Lusitania” is stylized as a documentary, and it is using intertitles in between the animated sequences as a narration to what is happening. It is an early example of virtual cinematography. McCay uses animation as a medium to recreate a historic event where a German submarine attacks an American ship. The animated short places a virtual camera in positions where it would be impossible for real camera to be. Most notably is the scene where the virtual camera is looking out from a window of the ship while it “records” the torpedo coming towards it, but also in more subtitle ways like when the submarine is passing underneath it.

Both these examples shows that animation from its early days has been used as a medium to construct “reality”, that is was used to depict elements, actions and effects that cinema could not. Noteworthy is that they are both following the narrative style of their live-action counterparts.

Virtual Camera and its narrative functions in animation and cinema

With the introduction of digital 3D animation in cinema and animation, there are no longer any limits to the freedom of camera movements and behaviors available to the director. Still constraints exist that keep the animated camera, with its possibilities to explore and construct any space and time, grounded to mimic the properties of its mechanical cousin.

This research is an attempt to examine how virtual cinematography is being used in animation and cinema in order to show its potential. I will do so by studying technological innovation and achievements in the use of narrative camera in both animation and cinema.

In “The Language of New Media” (Lev Manovich, 2001), Lev Manovich claims that “we do expect computer narratives to showcase new aesthetic possibilities that did not exist before digital computers” (Manovich, 237). He also claims that the directors refuse to give up the unique “cinema-effect” (Manovich, 310), an effect which according to Manovich depends upon three factors: the narrative form, the “reality” effect and the cinema’s architectural arrangement all working together.

This research will follow Manovich’s lead and discuss what role camera and constructed space plays in classic animation compared to modern animation and digital cinema.


Background

Lev Manovich, a professor at the Visual Arts Department at UCSD, gives us a critical insight into animation, cinema and compositing in his work “What is Digital Cinema?” (Lev Manovich, 1995), “Cinema and Digital Media” (Lev Manovich, 1996), “Assembling Reality: Myth of Computer Graphics” (Lev Manovich, 1999) and “The Language of New Media” (Lev Manovich, 2001).

In this research I am going to examine behavior of virtual camera in its relationship with the cinematic camera and constructed space. The term “virtual camera” is commonly used when describing the viewport of a computer generated 3D virtual space (“What is Digital Cinema?”, Lev Manovich,1995) and is used in related research papers when addressing topics surrounding simulated camera in computer animation (“Vanishing Point: Spatial Composition and the Virtual Camera”, Mike Jones, 2007).

“Virtual” is a term that is a reference to things that mimic their "real" equivalents, in this case a camera. So for the sake of this discussion I will use the term “virtual camera” as a broader definition than in those examples mentioned above, as this research will focus on simulated camera behavior and movement in constructed space in traditional 2D and digital 3D animation, as well as by motion control rigs and photo camera arrays. The similarity between these four different mediums is that they all simulate the functions of a cinematic camera, but in a constructed space.

I do not define the virtual camera as a thing by itself, but rather as the simulation of the properties of a real camera, in its relationship to its constructed environment. This phenomenon arises for examples when mimicking cinematic camera conventions through the viewport of a 3D renderer or when 160 photo cameras simulate one cinematic camera.

With the term “digital cinema” I refer to Lev Manovich’s definition. Manovich defines it as a cinematic production that has been digitally composited, where live-action material is composited with digital image processing, painting and animation.


Design of Work

In this research I will examine the specific grammars of virtual cinematography by studying technological innovation and achievements in the use of camera movements in both animation and cinema.

I will use the articles “What is Digital Cinema?” (Lev Manovich, 1991) and the book “The Language of New Media” (Lev Manovich, 2001) as primary sources for arguments about narrative grammars and cinematic reality. Both discuss the emergence of digital cinema and new media and their relationship with traditional mediums. Specifically I will take up Manovich’s discussion on the part virtual camera plays in this emergent context. I will also use “The Mouse Machine” (J.P. Telotte, 2008) as a secondary source in order to examine the history of animated cinema in order to find instances of the emergence of a virtual camera.

In order to examine the relationship between technical innovation and its relationship to changes in narrative grammar, I will look at the productions that introduced new narrative or technological innovation. I will attempt to use my findings to discuss how and where virtual cinematography is being used and how its properties are being explored, in order to show the differences and similarities between grammar of animated, virtual and cinematic camera.

Specifically I will examine the animated works “Gertie the Dinosaur” (Winsor McCay, 1914), “The Sinking of the Lusitania” (Winsor McCay, 1918), “Popeye the Sailor meets Sinbad the Sailor” (David Fleischer, 1936), “The Old Mill” (Wilfred Jackson, 1937), “Toy Story” (John Lassenter, 1995), “Tarzan” (Kevin Lima and Chris Buck, 1999) and “WALL-E” (Andrew Stanton, 2008). And I will be examine the cinematic productions “Love Me Tonight” (Rouben Mamoulian, 1932), “Star Wars: A New Hope” (George Lucas, 1977), “The Matrix” (Larry and Andy Wachowski, 1999) and “10,000 BC” (Roland Emmerich, 2008).

I will follow Lev Manovich’s lead and ask: What role does camera and constructed space play in classic animation compared to modern animation? How does the virtual camera in digital cinema represent a shift in the narrative grammar? In what ways has cinema introduced virtual spaces into its grammar?


Results


Cinema has always been about creating illusions, but at the same time they have hidden the illusions and special effects behind an accepted “reality” associated with traditional cinematic grammar.

In a way, little has changed since the 1930s. Unorthodox camera behaviors are being reserved for “special effects” shots, constructed virtual space is reserved for the “fantastical” and high-budget animated productions are still working hard to mimic the conventions of live-action cinema. This is because productions today still follow the same cinematic grammar as then.

Virtual camera as a part of cinema, still obeys these conventions. Where camera movements are liberated in space, which is the unique quality virtual camera contributes to cinema, it is used to promote an “out-of-this-world”-feel (“The Matrix) or a depiction of something “fantastical” (“Star Wars: A New Hope”).


Discussion

It is no longer the mechanical constructions available to the cinematographers, but the creativity and visions of the directors that set the boundaries of what is possible to depict in cinema. Pixar-director Andrew Stanton claims that since “The Lord of the Rings: The Fellowship of the Ring” (Peter Jackson, 2001), it is no longer any technical obstacles when it comes to creating photo realistic virtual worlds. But the cinematic language does not change in pace with the technological inventions. Digital cinema still depends on the “reality effect” the use of the traditional cinematic grammars brings to the screen.

Virtual cinematography as a part of digital cinema is superset to traditional cinematography in that aspect as with digital cinema, the directors have full artistic freedom to compose (and recompose) a scene. And as upcoming movies like “Avatar” (James Cameron, 2009) are using virtual camera to diffuse the borders between animation and live-action, I am certain that the use of the freedom of virtual camera in cinema and animation will broaden beyond the use as a special effect, and with that the grammars for what is accepted as screen reality will change.


Conclusion

With the introduction of digital 3D animation, there are no longer any limits to the freedom of virtual camera in digital cinema. But cinematic grammar keeps the virtual camera grounded to mimic the properties of its mechanical cousin, or to stand out proposing an “out-of-this-world”-feel. If current cinematic grammar is going to change drastically, cinema narrative has to be influenced by the narrative grammar of computer games, and thus introducing the spatial limitlessness of virtual camera as part of the convention for cinematic reality.

It will be interesting to see how virtual camera’s shift from post-production to the real-time recording in upcoming productions like “Avatar”, will contribute to create new narrative grammar specific for virtual camera in cinema and animation. As generations now grow up understanding and accepting new media languages, the virtual and computer generated will be more accepted as “reality”, without the need to hide behind old cinematic conventions to make them believable.

A further way to improve on this research will be to take the inclusion of music videos, anime, comic books and 3D cinema into to the study. It would also be interesting to compare the history camera narrative in computer games to the camera narrative in cinema and animation. A free roaming virtual camera is a natural part of the computer game narrative, so we have to look at current computer games to get a glimpse at the possibilities of future cinematic conventions for the virtual camera.