VFX Compositing - Montage, Collage and Decoupage

Lev Manovich observes in After Effects or Velvet Revolution (2006) (word *.doc) that compositing is premised on;

“the appearance of multiple media simultaneously in the same frame. Whether these media are openly juxtaposed or almost seamlessly blended together is less important than the fact of this co-presence itself… United within a common software environment, cinematography, animation, computer animation, special effects, graphic design and typography have come to form a new meta medium.” 

Manovich’s observation of the juxtaposition of media being less important than the coexistence of the media holds a great deal of truth in the context of the mechanics and logistics of how media elements are handled and treated in production. But the conceptual idea of juxtaposition, and what role such a long-standing artistic mode serves in compositing, is one worth digging into deeper in order to understand the aesthetic spatialities of compositing. A quotation from Sergei Eisenstein sets up some clarity on the idea of what juxtaposition is - the notion that “two film pieces of any kind, placed together, inevitably combine into a new concept, a new quality, arising out of the juxtaposition.” (1947. p4) But whilst this gives us some specificity for the idea in regard to film - conceptually a new thing arising from a combination of other things - it offers little to aid understanding of the techniques by which things are brought together in juxtaposition; how the technique shapes the result. Here there are potentially three terms that provide some insight on the differing methodologies by which visual juxtapositions are arrived at with each one presenting a technique and a spatial imperative that delivers a particular kind of juxtapositional result - Collage, Decoupage and Montage. Each of these three terms has a long and established history in artistic practice across different forms outside of cinema, and by definition the manner by which each is assembled provides a paradigm to view how compositing draws upon all three but doesn’t’ clearly or definitively tesselate with any.

We’ll start first with Collage. Certainly when we view a work like the title sequence from Carnivale we see elements of a distinctly traditional perception of a collage aesthetic. The word collage comes from a rather literal French term meaning ‘gluing’ and the tactile sensation of gluing pieces of media is certainly present in this work. Further, the very idea of gluing itself immediately conjures a sense of mixed media; that of gluing-on disparate objects to collectively make up a whole. However the spectrum of compositing (from a realism with intent to deceive, to an open and overt construction with distinct lack of indexical realism) means that collage isn’t as easy a fit with Avatar as it is with Carnivale. This is, of course, seeing collage as an Aesthetic rather than a Process; judging the terms’ suitability to describe the core of compositing based on how images rendered by the process appear. Obviously Avatar and the Carnivale title sequence do not share a visual aesthetic and thus if collage has value as a term to understand compositing it must be confined within process rather than aesthetic. The idea of collage - as a conceptual ‘gluing on’ - is central to the process of compositing though not necessarily the visual result. Here there is some common ground to be found between Avatar and Carnivale. The placement of a blue alien on the back of a flying lizard and set against a lush tropical world background is in essence a gluing of multiple elements one over the other. Each is (or is potentially) a different media type (CGI, photographic, matte painting, live-action, animation) glued together into a cohesive collage. Yet whilst the base idea of gluing-on disparate elements into a whole might hold weight, the problem of a well documented and rich history of collage-aesthetics in other art forms makes the use of the collage in regard to compositing problematic, or at the very least less than holistic. The core process of gluing on is difficult to separate from a particular aesthetic result and, as we have outlined, the fundamental power of compositing is the diverse spectrum of aesthetic results it can produce from a core set of tools and processes.

Decoupage as an artistic practice, and a particular method of creating visual juxtapositions, provides us a similar and yet distinct mode for thinking about compositing. As with collage, Decoupage is a world of French origin and stems directly from the term meaning to ‘cut out’. In traditional practice decoupage involved the cutting up of elements (usually paper images, photos, drawings, patterns and textures) and then pasting these down on a surface (most often furniture) and sealing them in place with layers of lacquer or vanish. By the nature of this process of cutting-up and pasting-down, decoupage invokes a kind of decorative aesthetic; an act of decorating something by juxtaposing layers of other visual elements on to and over it. As with collage, decoupage stands relevant to compositing as we can look at our case studies of Avatar and Carnivale and see in both some evidence of such a process at work. In both examples, the distinct visual elements might be viewed as lacquered onto a base - a use of layers to ‘decorate’ a scene through a decoupage of elements over the top of the scene’s base layer. Moreover the notion of selective cutting-out that is intrinsic to traditional decoupage holds value as a way of conceiving the chroma and luma keying processes of compositing whereby separations are enabled by removing particular colour or luminance values in the image; greenscreen chromakeying is effectively a kind of cutting-out. However decoupage is also problematic as an holistic reference for understanding compositing as it makes assumptions about hierarchy that are not consistent with the meta-medium of the compositing space. The core element of decoupage is not so much the ‘cutting-out’ as it is the ‘sticking-down’ and here decoupage carries the implicit implication of their exiting a primary base-layer or dominant medium onto which other, lesser/smaller media are to be stuck. Whilst we might look to Avatar and view the large-scale digital matte-paintings of the world of Pandora as serving that role of dominant base-layer onto which the cut-out / stuck-on elements of blue aliens and flying lizards will be decoupaged, this would fail to understand the process by which such composits are conceived and constructed. Greenscreen live-action or 3D CGI Composits need not begin with the background and indeed the background layer may well be the last element produced, let alone added, to the composit. And even then it retains its variability for change and substitution at any time during the process. By contrast decoupage implies an innate hierarchy - that the base-layer is a known and pre-existing entity onto which is juxtaposed the cut-out and stuck-on pieces. Decoupage certainly engages processes of juxtaposition but its assumption of a central fixed element onto which others are imposed is in opposition to the non-hierarchical nature of the compositing environment as Manovich observed.

This brings us to the third of the triumvirate, Montage, a term which has a rich and forceful history with a cinema specific context. Ostensibly Montage possesses much of the same juxtaposing nature of bringing together disparate elements as Collage and Decoupage. Conceptually embedded in Montage is a viewer engagement and reading beyond what is shown. As Eisenstein refers:

“The strength of montage resides in this, that it includes in the creative process the emotions and mind of the spectator. The spectator is compelled to proceed along that selfsame creative road that the author travelled in creating the image.” (1947 p32)

However the techniques and mechanics of how Montage is assembled evoke quite different considerations than those of Collage and Decoupage. Whilst the engagement of the viewer in a process of self-assembly to make meaning is consistent, the aesthetic outcomes are divergent and this has ramifications for notions of the intent for deception and acceptance of the image in compositing. Unlike Collage and Decoupage, Montage doesn’t derive form a particular act of craft-based manipulation. Where Collage means to ‘glue’ and Decoupage to ‘cut-up’, Montage having an origin meaning only to ‘put together’ lacks the same specificity and retains a more macro-conceptual air that theoretically may be applied to any number of more tactile and specific processes. But where we may see difference that marks Montage apart is in the expectations and use of media fragments themselves. The ‘Montage effect’ sits at the heart of the influential theoretical precepts of Eisenstein, Vertov, Kulesov, Pudovkin and which Eisenstein himself observed stemmed from the experience that; “While playing with pieces of film, they discovered a certain property in the toy which kept them astonished for a number of years.” (1947, p4) In a cinema specific context this casual observation indicates the dominant prerequisite for Montage to exist in delivering its audience-centric conceptual result - that Montage relies on the consistency of the media pieces brought together, that they are all of the same form. The famous Kulesov effect - whereby the substituting of an alternative second shot between the same first and third shots in a three-shot sequence, creates holistically different readings of the scene and it’s ‘story - indicates the reliance of a consistent media form in each of the components rather than the bringing together of different media types. In the most famous of Kulesov’s experiments a three-shot sequence of a neutral-faced actor looking off-screen is intercut first with a bowl of soup and then second with an injured girl. The result was observed as the actors ability to “convey with subtlety hunger in the first sequence and pity in the second.” (Ascher, S Pincus, E 1999. p347) For this Montage-effect to be wholly present in its traditional sense there must be a clear element of realism in the image joining - linear indexical coherence within the cinematic diegesis even if the joining is in defiance of the actual space-time of production. If the image of the bowl of soup were animated, CGI or if it was a graphic composited onto the footage of the man with the neutral expression, then we may certainly have juxtapositions (with collage or decoupage overtones, third-meanings arising from the joining but we wouldn’t have Montage as filmmakers have long understood it and used it to great effect. Montage as a practice doesn’t align with the ‘meta-medium’ concept of Manovich where by all media coexist because Montage relies on consistency and uniformity of the pieces being put together. The magic of the Montage effect that had Eisenstein and his contemporaries so astonished for years, is lost without such uniformity.

To return to the Composit, Montage doesn’t serve as a viable descriptor of the conceptual structure or process of compositing for three marked reasons; first the word itself already possess a specific branding and cannot be easily separated from its heritage in cinema studies and cinema practice to encompass a new visual structure and language. Second that Montage is by nature a linear assembly with its juxtapositions predicated on sequence and order rather than coexistence and simultaneity. And third, that Montage avoids the common element that unifies all the technical process of compositing and which dictates the aesthetic results (either realist or formalist) of composited media, Transparency. Indeed whilst all three of Collage, Decoupage and Montage create juxtapositions and drive the proactive engagement of the viewer in constructing for themselves third-meanings from the connection of images, none of the three by-nature possess a natural presence of Transparency. The very core of the compositing process is the generation, management and manipulation of Transparency - the Alpha channel. It is this element (at once both mechanical tool and conceptual idea) that separates out ‘Composit’ as an artistic term of its own unique value from it’s siblings of Collage, Montage and Decoupage. Compositing is, in effect, a unilateral simulation of parallel forms in a singular space with each form bringing its own language tenets to bare on the visual result.

To distill the fundamentals of these three traditional constructs we can view them by how they bring media together. Collage is different media assembled together to form a gestalt without a dominant media. Decoupage is different media adjoined to a single, dominant base media. And Montage is the drawing together of same media in a linear assembly. The simplicity of these descriptors opens up discussion around how each form handles the individuality of each media piece, how recognisable the individual parts are within the construct of the whole and where the viewer’s experience of juxtaposition resides?

Collage : Individual elements identifiable Individually but seen Simultaneously - juxtaposition by spatially adjacent proximity.

Montage : Individual elements identifiable Individually and seen Individually - juxtaposition by sequence and temporal arrangement.

Decoupage : Individual elements identifiable Holistically and seen Simultaneously as belonging to a dominant - juxtaposition by spatial decoration.

These definitions lead us to an understanding of Compositing which positions it as distinctly and divergently separate and apart from traditional methods of visual juxtaposition.

Compositing : Individual elements that may be neither identifiable individually nor seen individually and which are perceived both sequentially and simultaneously - juxtaposition by visual and temporal arrangement with in a uniformity of space.

The difference between the spatiality of collage arrangement and the spatiality of compositing is that transparency and layers makes the identification of independent parts, which is fundamental to the idea of reading juxtaposition, difficult to discern. So the spatiality of compositing is of unity and singularity rather than the variation of collage. Transparency by nature theoretically makes infinite the number of elements that may coexist in a single temporal moment. 

+++

Bonus podcast of earlier lecture on this topic….

COLLAGE, MONTAGE AND JUXTAPOSITION
The space of transparency - motion graphics, visual fx and compositing
Music videos, TV title sequences, greenscreen shots, digital animation and composited 3D spaces - the language of contemporary cinema has never required such diverse mechanics. And at the heart of all theses processes is Transparency and the construction of seemless visual layers. This presentation, given at the University of NSW May 26th 2009, teases at the unique structures of motion graphics production and the ramifications of multi-layer visual production on traditional notions of montage and collage.
Download audio *.mp3

___

Manovich, L., 2006. After Effects, or Velvet Revolution in Modern Culture. Part 1, Available at link

Eisenstein, S., 1947. The Film Sense 1975 ed. Jay Leyda - Translation, San Diego: Harcourt Brace.