Our ideas evolved as the project developed. We began thinking about the rules of composition and how artists have, through years of trial and error, determined the compositions that attract our eye in a certain way and allow them to effectively manipulate our attention as we view their work. We considered the effects on visual attention resulting from use of the golden section, the rules of thirds, apex/pyramidal composition as well as tricks of the eye as seen in Monet’s 1872 'Impression: Sunrise', and Escher’s playful yet frightening confusions created from distorting perceptions. We then applied cutting edge technology to explore what really does happen to the eye when faced with such images.

Initially we began with well-known pieces – some examples are displayed on this site. These include Munch’s 'The Scream', Da Vinci’s 'Monet Lisa' and 'Vitruvian Man', Manet’s 'Olympia' and Vermeer’s 'Girl with a Pearl Earring'. We also looked at 'The Creation of Adam' from the Sistine Chapel, painted by Michelangelo Buonarroti. Due to the well-known nature of the works we chose, it is worth noting that none of the works were naïve to these images. We viewed each image for 30 seconds, a duration we determined (through careful testing) to be long enough to view the entire picture but short enough to determine the locations of the main fixations (points where the eye stops and dwells) without too many 'repeats' or returns to the same spot.

From this data we were able to see where the eye is drawn to by specific features of the image, and infer how the artist attempts to manipulate the viewer’s attention. An interesting angle which we have yet to pursue, but continue to ponder, is the differences in how 'The Artist' and 'The Scientist' initially viewed the art. A related question also arose: does repeated exposure to the stimulus alter the salience of the techniques used by the artist and perhaps allow 'freer' interaction with the piece unconstrained by the initial 'attention grabbing' features? This was a digression we have yet to explore - a project for the future (or a thesis for a budding artist/scientist?)

In Phase 2 we trained our eyes to make a mark. In the same way a painter would repeat and test out different strokes and techniques, so we spent many hours in the lab training our eyes. We marked (effectively tracing with our eyes) everything from wine bottles to Bhuddas via the Twin Towers and Sydney’s Opera House as well as iconographic images such as dollars, pounds and euros, and religious symbols. A limited selection of these prototypes are available for purchase and can be customized to your preference. Please contact us at this email if you would like more information: Tamara Russell/Hamish Ta-Mé

This training phase was exhausting. Conscious controlled eye movements are the domain of the brain’s frontal lobes – we certainly worked those lobes! Day-to-day variations in our eye-marking ability were evident; Hamish could tell if Tamara had been out the night before and Tamara could tell if Hamish’s newborn had kept him awake into the early hours.

We tried different techniques to make the mark – multiple, short fixations versus sweeping, controlled saccades. With the former, it was easier to delineate the content of the picture, but the latter created a more dramatic final product and is more akin to the 'drawing' process in traditional art. We also noticed we adapted slightly different approaches to our technique (for example starting in different locations or having a particular internal model of how we were going to 'draw' before starting). This may be a manifestation of the 'Artist' and 'Scientist' mindsets. The two-person approach was crucial as when marking the image the only clue to where you have already marked is internal (in your mind’s eye), there is no external feedback telling you where on the screen you have already 'visited'. The collaborator, sat at the main EyeLink PC, could see where you had visited and alert you to parts that had been missed or needed further definition. As we refined our technique however, this requirement became less and less.

A key point that needs to be made here is that these images are not 'free drawing' with our eyes but rather stimulus-driven. We believe that with continued training free drawing would be possible although this is beyond the scope of this current project. The lack of feedback (described above) was the most serious technical challenge to this attempt at free drawing. Real-time feedback from the 'artist' PC could allow free drawing to become a reality. One application for this might be to allow individuals who are paralysed (but with eye movement) to create their own 'bioart'. Biofeedback systems for other physiological measurements (such as galvanic skin response – how much you are sweating) are currently in existence, so this is not so far-fetched.

In Phase 3 we went back to the original artworks and this time used our trained eyes to make a tracing of the original. Typically the duration for each tracing was 180 seconds and the majority of pieces were done in one sitting.
In Phase 3 we went back to the original artworks and this time used our trained eyes to make a tracing of the original. Typically the duration for each tracing was 180 seconds and the majority of pieces were done in one sitting. In Phase 3 we went back to the original artworks and this time used our trained eyes to make a tracing of the original. Typically the duration for each tracing was 180 seconds and the majority of pieces were done in one sitting. In Phase 3 we went back to the original artworks and this time used our trained eyes to make a tracing of the original. Typically the duration for each tracing was 180 seconds and the majority of pieces were done in one sitting. In Phase 3 we went back to the original artworks and this time used our trained eyes to make a tracing of the original. Typically the duration for each tracing was 180 seconds and the majority of pieces were done in one sitting. In Phase 3 we went back to the original artworks and this time used our trained eyes to make a tracing of the original. Typically the duration for each tracing was 180 seconds and the majority of pieces were done in one sitting. In Phase 3 we went back to the original artworks and this time used our trained eyes to make a tracing of the original. Typically the duration for each tracing was 180 seconds and the majority of pieces were done in one sitting. In Phase 3 we went back to the original artworks and this time used our trained eyes to make a tracing of the original. Typically the duration for each tracing was 180 seconds and the majority of pieces were done in one sitting. In Phase 3 we went back to the original artworks and this time used our trained eyes to make a tracing of the original. Typically the duration for each tracing was 180 seconds and the majority of pieces were done in one sitting.

We are now in post production Phase 4. We have taken our initial eye movement responses to each image in Phase 1 in order to determine where our attention was drawn on the first viewing of the work. This sequence of eye movements represents a combination of the artist’s intentional direction and redirection of our visual attention, as well as the individual approach we each brought to the work. The fixations from this first viewing were then overlaid on the Phase 3 traces of the work. Remember in Phase 3 we used predominantly saccades to trace out the details of the image, in essence using our blindness to make a mark that can be observed by you.

From the start we were aware of the stark juxtaposition between the creative process of these classic works and our high-tech laboratory conditions. However we had predominantly in mind that the final product should have a classic 'art' feel, despite the high-tech twist. Therefore, the final artworks are presented on high quality paper and printed to the finest standards, as we appropriate the language and presentation of classical art for our work.