Rebody : Consideration Figure

Input data for Rebody consists of the motion-tracked data of the body of a dancer. Approaching this data from the point of view of 'figure', it is clear that a number of elements need consideration.


More information on my use of ‘figure’ can be found here.



The human body


When the dancer walks, we are very quick to see this in the coordinated movements of the data points. Initially, this produces the effect of visualisation rather than figuration. To move Rebody into the direction of the latter, the human Gestalt that we see in the data needs to be suspended.


This is also a question of choreography.


The movements that underlie Rebody are not purpose made; they were part of a different project, Bodyscapes, in which the dancer by moving in space could create sounds that, at the same time, she responded to. This setup suggested bigger movements of the body as well as ‘walks’ across the stage. 


More information on Bodyscapes can be found here.




In order to focus on the figural aspects, one of the first decisions was to ‘pin’ the data at the centre of the ‘body’ i.e. the recorded data of the thorax. Setting the thorax as origin of the coordinate system means that movement is only ever relative to this point. (The thorax itself is not used for the drawing.)


This pinning corresponds to the way I usually place figures in the centre with slightly more empty space around it than one might expect. This resonates with historical scientific illustrations but also with the tidy situation in a lab. As I say in Paris (2008: 81), figural drawings ‘look like they have been made under a microscope.’ Dealing with moving data, this aspect perhaps produces negative undertones: a living animal that is fixed on a plate for further studies in the lab.


Reduction of data


The reduction of data points from 17 to 12 has three implications.


(1) Abstraction: less data points give a less detailed image of a body. While the data points that were left out (lclavicle, rclavicle, pelvis, lfemur, rfemur) are not crucial for a reconstruction of the human body – they are often in the proximity other data points – a tendency towards the abstract has to be noted. It would have been more extreme had we chosen to just pick one further point, in which case Rebody would have resulted in a ‘dancing’ line.


(2) From pattern to figure: a drawing can quickly become ‘crowded’. There is a moment where our seeing habits switch, and we become overwhelmed by the amount of data that we need to process. In turn, this reduces our attention per point i.e. we cannot simultaneously keep all points in focus: the drawing ‘blurs’, and we start to see a moving pattern. This may not be unrelated to research from the middle of the 20th century, which, reported by people such as Abraham Moles and Herbert W. Franke, estimates that a human can on average process a maximum of 16 bits per second of new information. Even if it is not necessary to infer aesthetic criteria from this as was fashionable at the time, one may still want to say that ‘figures’ address the perceptual apparatus in specific ways remaining within its processing power.


(3) Rebalancing: The motion-tracking apparatus displays a bias towards the upper body part. This, and the extraordinary length of the femur create an imbalance in terms of point distribution. Such an imbalance makes it almost impossible to produce a sense that the figure may hover and that it may not already be directed.


Start point


The reduction of data points makes the choice of start point for the drawing less of an issue. However, even with less points, an erect body with arms up and a start point in the upper body half, will in most likelihood not be completely drawn since the drawing 'exhausts' itself in the upper part, which is caused by the drawing algorithm. (See section: Data Processing) Starting with the right foot, on the other hand will support the overall balance of the figure and prevent sudden reductions of the figure’s complexity when not all data points are used for the drawing.


Reduction of speed


The reduction of the speed of movement (from 120 to 12.5 frames per second) also reduces the amount of information to be processed. Given that it is a less ‘natural’ speed of the human body, it adds a further layer of abstraction to the drawing. The quicker the data points move, the more body becomes visible in the figure. Independent of the figure, the speed reduction has grave consequences for the sonic part of Rebody: the sample rate of the motion-tracking apparatus becomes clearly audible and part of the texture of the piece as a whole.


Sample selection


Still, in spite of these adjustments, looking at the recorded data it was clear that one was not automatically given a moving figure. In most cases, in particular when the dancer’s body was upright and striding, the figural suspension of the human body collapsed into its visualisation. Thus, Rebody needed to use only those samples in which the figure ‘worked’. This is one of the reasons why the final piece consists of samples rather than a continuous sequence. Having Rebody consist of samples adds further weight to the above mentioned ‘scientific’ style; naturally, it also resonates with serial art: a thought that is stretched across a number of variations, which no single image alone is able to represent.


While serialisation would certainly have worked in the case of Rebody, the material suggested two further directions that we did not want to leave out.




The first additional aspect has to do with the motion-capture apparatus, or rather glitches in it. Motion tracking uses a number of cameras arranged around the object to be tracked. If the object in a particular (dance) move prevents a certain number of cameras from recording sufficient tracking points, the movement of the resulting three dimensional model becomes less and less smooth. In other words, just like the sample rate that impacts heavily on the sounds, the trembles and jitters of the drawing affect the figure. In effect, when they occur, we are made more aware of the underlying data processing and the virtuality of any movement, body or figure that we perceive. These interruptions further suspend and thus challenge the figure.




The second additional aspect concerns the shape of the figure. As mentioned above, the long and thin shape of the standing body is easily recognisable giving the figure a strong directionality. When looking at the samples it became clear that, overall, Rebody should not consist of too many of these instances so that when an upright figure appears it would be read as one possible, but not the dominant orientation of the figure.


To support this, we experimented with the circ-function, the only operation in Rebody that moves data points. The circ-function projects all points onto a circle drawn around the centre of the figure (thorax). It can be fractionally activated so that points are moved towards the circle without being flattened onto it (the figure is ‘blown up’). With the perfect circle, the circ-function suggests an extreme other to the body – an eternal shape – which also suggest a body’s horizon or its aura, even. Naturally, Leonardo da Vinci’s Vitruvian Man comes to mind.


In the case of full-circ figures (the beginning of Rebody), it is more the border and less the centre that appears to be pinned down although the pelvis remains the point of reference. Movement of the figure becomes almost completely internal, however.

See section: Samle Matrix for a comparison of circ settings.




With the help of the circ-function as well as the variations of the sample in terms of shape, a narrative across Rebody could be assembled. The narrative chiefly consists of three parts:


(1) Dominant circ-function at the beginning and a long first scene that opens the piece. Since through the circ-function the body is most suspended, movements in the figures may remind a viewer of the kind of coordinations inherent in the human body. Thus, the first section anticipates a body (that never quite arrives).


(2) In the second section the image of the body is allowed to shine through. Certain gestures can clearly be read as belonging to a human body. Most vertical figures are in this section that often also exhibit a sense of three-dimensionality.


(3) The last section consist of a return to round shapes reminiscent of the beginning yet completely different. There is very little human body visible suggesting some kind regression of the figure towards amoeba-like life forms. With the round shape and slow movements, this section also provides some form of self-containment of Rebody as a whole.



Rebody as figure


As the above considerations indicate, a fairly precise idea of figure has been sought in the data collected from Bodyscapes


There are striking similarities with figures in previous works, which have been constructed in very similar ways, the drawings for Paris (2008) are an example. Yet, each figure seems to have ‘a life of its own’ that feels connected to its root in the data.


It is difficult if not impossible to speculate how figures allow us to see the data differently. Transpositions, the collaboration that followed Rebody, is set up to test this further by seeking figures in very different, scientific data sets. In Transpositions, scientists are enlisted to help better understand more precisely what figures may have to say about the data they ‘represent’.


Schwab, M., 2008. Paris, London: Copy Press.