The Development of the Covertor Concept

Between 1^st October 2003 and 9^th January 2004, Russell Richards worked on the creation of a new type of digital art creation tool named 'Covertor'.

 

Digital art practice can be stimulated by a wide variety of components. In the case of the covertor the stimulus in the first instance was not the reinterpretation of a dense concept, it was not the result of happenstance or the representation of a media form within the digital or an emulation of a real world object. The source for the Covertor concept was a facility within Macromedia Director that enables the author of an application to choose to 'animate in the background'. This facility means that an application can be created that sits on the desktop and is active at all times. A classic use for this would be a music player app that continued to play music whether it was 'in focus' or not.

 

This facility was conjoined with a concept that Richards had been teasing away at for a few years i.e. how minimal interactivity could create maximum effect. His previous venture into this area had been 'cairn' an on-line Macromedia Director Shockwave. This application required nothing more than the user returning to the site to see that their  presence on the site had made a difference. In the case of the cairn each time a user returned to the site an additional stone was added to a pile a la adding stones to a cairn on a hillside i.e. as a statement that 'I have been there'. The bringing together of these two elements (animate in the background and content production that relies specifically on the user) enabled the development of an app that could interact with the user while in the background. In this case the mode of interaction chosen was that of the creation of a digital art piece. Macromedia Director was used as this package is Richards' 'weapon of choice' for making digital arts works. This is significant in itself as Director is known predominantly as a CD-ROM creation tool. However, Director can be utilised to create virtually anything including digital art works. At time of writing we do not know of anyone else using Director for such a purpose.

 

Fig. 1

 

The initial specification for the Covertor was as follows:

1. That it should be an application that runs behind other applications.
2. That it should respond to mouse clicks or moves.
3. That a 3D sculpture is produced.
4. That at the end of the work session the user could screen grab their art work and save to disk/print it out.

The mode of operation here was to run the Covertor in the background while using Microsoft Word or Safari browser or Excel Spreadsheet etc. then sense the mouse activity back in the Covertor as the user performed their functional tasks with their software. Richards quickly realised that it was not possible to sense a mouse click in one application in the Covertor beneath. However, the Covertor was able to sense mouse movement ('mouse-tracking'). At the end of the session the user will have created a random art form out of their purposive movements across the screen. The earliest version, shown in Fig. 1 above,  involved a line being drawn from a central point with a randomised colouring effect. This initial version was simply to prove that it could work at all. The result was interesting, however, Richards as mentioned above, was interested in incorporating additional dimensions into the application. Rather than moving into 3D (a facility available within Macromedia Director i.e. Shockwave 3D) Richards decided to use degrees of opacity to create inferred depth.

 

Fig. 2

 

The next investigation utilised a different method of 'mouse-triggering' here a graphic was rotated around the mouse position to create washes of colour over large areas of the window. Fig 3. Richards was pleased with the subtlety of the outcome.

 

Fig. 3

 

However, he then made a further departure from the brief by producing a version of the covertor that did not track with the mouse at all. Here a lateral investigation into the capabilities of rotation of graphics took place. In this way the outcome of happenstance and tentative investigations can be incorporated back into the main project. A Covertor that does not allow mouse-tracking is not a Covertor. Yet the reapplication of one aspect of the Covertor, isolating it from the remit enables, a close examination of its properties. Instead of mouse-tracking one of two modes were employed to situate the graphical elements in the window. The simplest method of situating was to centre the graphic and allow rotation Fig. 4 below.

 

Fig. 4

 

This first iteration produced an output that seemed to be an homage to Len Lye the film maker. Spinning circles of light coalesce then overwrite each other until near blackness is left. The second iteration involved the graphic being controlled through code by randomising the position of the graphic within the window. However, the simplicity of the effect was completely disrupted by this randomisation and so centring was used instead. The utility of randomisation was placed in an archive for possible use later on in the project. Centring was not adopted as primary part of the specification for the Covertor. Indeed, as mentioned above, mouse-tracking not centring is a determinate property in the construction of a Covertor. Centring had proved useful in delivering one new form of dynamic digital art form. However, if further investigation was to occur then there had to be a return to the core principles of the Covertor. But this did not stop Richards exploring yet another approach loosely built around some of the Covertor's properties. Richards has had a sustained interest in alternative cursor forms. Before returning to the creation of other forms of Covertor Richards took some time (2 days) to examine whether mouse-tracking and rotation and layering could be utilised in the creation of innovative cursors. The result can be seen in Fig 5. Richards considered this to be unsuccessful and put the idea to one-side. Further short term experiments were conducted in alternative recursions of particular shapes Fig 6 and Fig 7 show two examples of these experiments.

 

Fig. 5

 

 

Fig. 6

Fig. 7

 

These experiments were interesting but none of them created a new form of dynamic digital art that Richards was particularly pleased about. He made the decision to return to the main specifications of the Covertor. This time the mouse-tracking employed utilised a predefined circular motion instead of a rotation around a point. The resulting output produced an extremely dynamic relationship between the position of the cursor and the graphic. Richards was particularly pleased with the result. The output from these versions was quite 'Russian Constructivist' in style with strong angular lines and vortices. See fig. 8 below.

 

Fig. 8

 

This step represented a major point in the Covertor development. The inter-relationship between the mouse and the graphic display enabled subtle layering and geometric shapes to be created successively. In this way, the specifications of the Covertor (bar the 3D sculpture component) could be implemented in a manner that Richards found extremely satisfying. This could have been an end to the project with the 3D sculpture component left in abeyance. However, Richards was sure that a 3D component could work within the remit and he persevered with the idea. The immediate response was the incorporation of small graphic elements that prescribed the same circular motion as earlier iterations but simply created successive loops on the screen. See fig. 9 . The aim here was to create a layering of graphics that were additive. Over time a 3D sculpture could be produced. However, the result was crude and messy in a way that Richards did not approve of.

Fig. 9 This approach was left and Richards then reassessed the previous non-mouse-tracking iterations in terms of the rotating code attached. This new approach spawned a series of Covertors that operated at a different pace to the previous versions. Indeed, the next fifteen Covertors utilised a comparatively slow growth process. The first of these Covertors was graphically 2D. However it was generated from a bitmap that was web-like in construction and this allowed for a mix of overlaying and revealing over time. See Fig. 10 below.

 

Fig. 10

 

This slow overlaying was combined with other code effects to create a further two versions i.e. Fig. 11 utilising a stronger pinching effect and Fig 12. removing the rotation from the array of effects.

 

Fig. 11

 

Fig. 12

 

Although these effects produced some interesting outcomes Richards reference point for the Covertors had become the subtlety that could be achieved i.e. Fig 8 above. He attempted to reincorporate some of that subtlety by reducing the number of animated elements and remove the randomisation of the colour. Furthermore, Richards counterpointed the animations by having one graphic rotating clockwise and one anti-clockwise. The result shown in Fig. 13 combined the simplicity of the colour use and repetitive movement with a large amount of variety possible simply through the change of starting positions for each rotation. Sometimes the one graphic would overwrite the other and sometimes vice versa. Sometimes the two graphics would cancel each other out sometimes they would pirouette around each other. This iteration moved at the same considered pace but unlike the previous iterations (Figs. 10, 11, 12) its output was unpredictable and thus much more stimulating.

Fig. 13

 

Fig. 14

 

Richards then made an intuitive leap. It is possible to use a variety of 'ink' effects in Macromedia Director. These effects are, in some cases, similar to those available in Adobe Photoshop layers. However, if they are applied to animated graphics then they add further dynamic attributes to code-oriented effects. By applying the 'subtract' ink to the two rotating graphics and screen grabbing the output at the end of every full rotation a remarkable change could be created in the Covertor. See Fig. 14 above.

 

The output was in some ways reminiscent of 'Magic Eye' pictures or sand paintings. However, here the patterns created were caused by the randomising effects caused by a mix of ink and rotation. This output seemed to coalesce and in the advanced versions continued indefinitely to create new forms and patterns.  In some cases, Richards can use his experience of nearly 10 years with Macromedia Director to predict what a given piece of code or an effect may have on a project. This was not one of those occasions. There is certainly more that could be done with this 'sand painting' mode of the Covertor. However, in terms of the core of the Covertor specifications it did not satisfy all the criteria so it was left to one side. Richards made one more foray back into the previous iteration (i.e. Fig 12) in order to experiment with contrasting colours and ink effects. The result of which was pleasing to Richards as he was able to relate austere dark strands with almost fluorescent gold lights. See Fig. 15 below. However, Richards considered that this particular series of renditions had run their course.

 

Fig. 15

 

Richards then sought to redeploy the use of subtle changes in blending/opacity that he had been able to utilise to his satisfaction in early iterations of the Covertor. The thought process here was that the slower pace of the animations latterly employed could create some subtle effects if allied to slow blend/opacity variations. These variations were working down to .01% opacity. In this way the building up of successive layers could be combined with a slower paced development in contrast to previous Covertors layers that had been obscured by frenetic animations and the 100% opacity of graphics. The next 6 iterations of the Covertor used this combination of effects. The first of these created some 'Turneresque' outputs. See Fig. 16.

 

Fig. 16

 

This was operating at a qualitatively different level of subtlety compared with previous versions. The mix of washes and light intensities gave a form of three dimensionality to the output. This was light creating landscape. The Richards made a further intuitive leap. What if one of the 'sand paintings' was utilised as the original graphic for this latest iteration? The result as shown in Fig. 17 was remarkable.

 

Fig. 17

 

Colour washes were intermixed with what looked like scars pulsating horizontally across the screen. The effect was sublime. Richards has extracted this version of the Covertor out from the array and intends to exhibit it separately under the title 'rent'. Further versions of this iteration were created using different graphical elements examples are shown in Fig. 18 and Fig. 19. The later version created phased pulses across the screen during a session.

 

Fig. 18

 

Fig. 19

 

As previously these investigations were interesting and indeed fruitful but the Covertor specifications were once more being sidelined. Richards' response to this was again to return to fundamentals and attempt to create objects that incorporated a 3D component. Two versions were created that displayed discrete graphical elements again layered with varying degrees of opacity. See Fig 20 and Fig 21.

 

Fig. 20

 

Fig. 21

 

These were interesting but Richards became frustrated with the inability of the code so far employed to enable all of the parts of the specification to be implemented. Encouraged by work by one of his full-time MA Interactive Production students, Richards finally went to Macromedia Director's Shockwave 3D facility. To avoid the intricacies of creating 3D shapes on the fly, he used a simple piece of code that could generate 3D text in an environment. This was combined with the library behaviours for rotating a 3D object around an axis and the code to generate new objects at regular intervals. Mouse-tracking was incorporated by setting up a relationship between the mouse position relative to the centre of the screen and the size of character produced a selection of outputs are shown below. Figs 22, 23, 24, 25.

 

Fig. 22

 

Fig. 23

 

Fig. 24

 

Fig. 25

 

These experiments offered Richards great hope that it was possible for Shockwave 3D facilitate the satisfaction of all the specifications for the Covertor i.e.

1. That it should be an application that runs behind other applications.
2. That it should respond to mouse clicks or moves.
3. That a 3D sculpture is produced.
4. That at the end of the work session the user could screen grab their art work and save to disk/print it out.

However, there was a tension between the generation of 3D graphics in a session that had built up a 3D sculpture and the implications for the computer's processor as more and more graphics and their motions had to be calculated. On Richards' computer test bed once 500 3D objects had been produced the computer would freeze. This presented a awkward contradiction: for a Covertor to  work it had to create an art work 'in the background' but no fixed time could be allocated precisely because the user was busy working in another programme. For the Covertor to work covertly it should not call on the user to have to intervene in the process (in order to stop the computer crashing). Richards initial response was to create a Covertor that deleted 3D objects once an optimum number had been produced. This solved the problem of crashing but produced a Covertor that did not accrue an art object over time but constantly created art objects in 50 object bursts - therefore it had to be watched almost constantly. Some interesting effects were produced that were somewhat similar to flash-style eye candy where a user moves their cursor to create 2D or 3D graphics on the screen. However, this was in fact the antithesis of what a Covertor was supposed to be as laid down in the above specifications. After a few days reflection Richards designed a Covertor that utilised the screen grabbing facility as previously applied to the 'sand painting' Covertors. The theory went as follows: If a screen grab of 50 3D objects could be captured and then placed as a background then a further 50 3D objects could then be created on top of that and so on. This would remove the processor issues and enable a composite 3D sculpture to be created during an entire working session. That was the theory and Richards was able to create a version that would work from an exe application.

 

 

Fig. 26

 

This grabbed scene, Fig. 26, was captured at a lower resolution so that when placed through code as a new background it appears fuzzy. Richards did not particularly like this lack of definition but some of MA Interactive  Production students thought that it encouraged the feeling of depth. This was a version that, for the first time, actually achieved  all the specifications for a Covertor as laid down. However, the code and utilities employed to produce this version do not work in a browser. The aim had been to include the Covertor in the Hidrazone as an on-line digital artwork. Yet,  at time of writing,  the purest form of Covertor can only be viewed/used  off-line.  This method of artistic creation of successive iteration and regression into previous versions has enabled the production of a wide variety of artworks. Here randomisation has been utilised as an active component of the creation process i.e. the direction that a given aspect of the Covertor creation process might take was not pre-ordained. Several of the more interesting versions of the Covertor were the result of intuitive jumps into the unknown and indeed the result of happenstance.  But how much more interesting to take this approach than to discard all but those pure forms of Covertor. Indeed, if a more systematic approach had been taken it may have quickly become apparent that, in fact,  an on-line Covertor satisfying all the specifications may not be feasible and the project would have been terminated before anything was created!