This eyetracking business needs further (over)-simplification. And re-mystification.
Because I can move nothing but my eyes, we use my eyes to move my computer mouse. This works quite well till it comes to gesture.
Here’s how it works:
An infrared beam illuminates my face and causes my eyes to reflect brightly, like a cat’s or wolf’s, and the angle of this reflection is carefully calculated against known points on the computer screen.
In this way, the computer knows where i’m looking and places the cursor there.
After a moment, a magnifying glass zooms in on the cursor area enabling me to place the cursor with near pixel accuracy. After another moment the magnifier snaps out of the way and the click is made.
This is eyetracking. As practiced in 2013.
Of course, there’s more.
The whole of art history gets a lovely kink in it right here. I expect you may have noticed it by now. It’s worth taking a minute to explore.
When the gaze actually becomes the line, an ancient polymorphous perversion has been fulfilled – the eye without mediation – as the line maker.
One step closer to some more unnamable secret dream of thought visualization, telepathy or dream recording. Furthermoreover, one appears to have reversed the direction in which eyes are designed to work – one is courting Faustian retribution. One is almost reversing the influence, certainly inverting and feeding back the act of looking, mechanistically nesting the concept of voyeurism, and actually physically reversing the flow of photons; the eyes are doing the emission.
Another avenue for the human will to master creation…clearly a signpost pointing to another enthusiastic joyride of hubris… Ok, that’s perhaps just overkill… But a complicated and seemingly miraculous alchemy, clearly against the ‘divine’ order. Or is it?
It may interest you to know, that this was actually the prevailing theory of sight for thousands of years — that sight emanated from the eye, emitting rays like, (insert favorite monster/villain here)
And not just any thousands of years, some of our favorite philosophers were quite convinced: from ‘the Fifth Century BCE, Empedocles concluded that Aphrodite lit a fire in the eye which shone out from the eye making sight possible’. Sometime later, around 400 BCE, Plato was still cool with that. A hundred years after that, Euclid, in his ‘Optica’ – by now definitely showing signs of suspicion – stopped just short of declaring it foolishness. Two hundred-fifty years later finds Lucretius: still down. Ptolemy (c. 2nd century) was an adherent to the ’emission’ theory of vision… anyway, so it is written.
The first sign of correctness came around the 11th Century when the Actual Experiments of the Arabian physicist, Ibn al-Haytham the Overconfident’ led him to return to the facts in his ‘Book of Optics’. Essentially, people were still bickering about it ’til the speed of light (and thereby the light cone of causality and logically the mechanisms of sight) was proven.
Frighteningly, Winer et al. (2002) have evidence that as many as 50% of American college students still believe in emission theory.
It’s a wonder three point classical perspective ever got started based on such a shifty foundation, and that seem a great hint at the human difference – unlike other popular organic minds, when we bumped into it, we really ran with the transformative abstraction gimmick.
We busted into information-space like it was ours all along. And we have yet to meet inhabitants (which there surely are :- ) or encounter any resistance (besides our own amphibian-like clumsiness) it’s only a matter of swiftly decreasing time before we officially colonize info-space and move in.
Eyetracking, (in it’s present form) does not possess depth coordinates, no z axis if you will, and naturally makes the picture plane flat, flat flat, even possibly slightly concave. (Hmmm, might have to do something with that, eh?)
In this peculiar case, the eye does indeed project the line, so in one stroke – a severe repudiation of the tenets of classical art is accomplished and the rhetoric of abstract expression is ratified with little fuss. To say it once again, these are the ramifications of eyetracking, not personal opinions, I rather like perspective.
Since you are working with a virtual beam, stepping back to observe the gestalt of the image is a major production, a deliberately passive operation, totally divorced from the making. This of could be good, because you tend to see things anew a little faster than a hand-painter might.
Even masters of line and balance of forms, I dare not speak their names, use their peripheral vision to judge their progress and effect as they paint.
This is the main aberration of eyetracking for drawing and painting. Even the most pure of action painters or sketch artists devoted to line isn’t working with laser eyes, for the slightest attention to peripheral vision in eyetracking results in deviation of the gesture. What i’m saying is, it’s hard.
Willpower and intention/attention is a palpable commodity in eyetracking and intention focused to the point of total exclusion.
This is a severe misuse of the eye, an insult to aeons of development, rather like using your laptop to rake leaves.
The sophisticated organic in-eye processing certainly still comes in handy, along with muscle memory and minds-eye mapping, but the most exquisite arrangement of peripheral luminance overemphasis and tracking and motion sensing, (which works in the most convoluted of biochemical schemes, that i won’t get into here) along with an eternity’s worth of symbolic and field processing – is whittled away to a fine point.
The classic quote: ‘the (computer) mouse is probably the narrowest straw you could try and suck all of human expression through.” — Joy Mountford
Comes to mind…
Of course, the computer is such a plastic medium, a multitude of workarounds exist, like digital audio hyper-instruments before them, one can utilize editable vector graphics to fine tune eye ‘gestures’ or dynamic compression to smooth out corners in real time, or distribute color and brush/line shape according to the length or speed of your eyestroke, or according to a ‘reference’ image or even use a ‘seed image’ analysis to ‘steal’ colors, shapes, and abstract and recombine their features separately. i use all of these tricks and more.
Just to function, the eyetracking unit/software is already making fairly massive modifications to the rendered gaze data, pruning jitter, involuntary sacading (sweeping eye movements, unconscious micro-scans) blinks, and reflection noise near the target infrared spectrum, etc,
But further modifications could be made.
Dwell, the time it takes the eyetracker to decide you ‘really’ want to click where you’re staring, (in my case set to impatient – is certainly an area for improvement, a modal, or context sensitive dwell might be useful.
Getting off the subject, but hopefully fascinatingly so, the nervous system affords several other points of access for opportunities to sample values to be used for parameter modulation. If combined, perhaps a truly subtle (though devilishly hard to master) interface suite could be fashioned. I imagined such things as a child, studied them in college – it’s here now and that’s what i want to work on next.