Last night’s coding session (when you’re too poor to buy a Kinect)
For some time now, I have been fixated on initiating this physical computing project that involves a swarm-like interactive animation, and an ultrasound sensor. Last night, I finally managed to start working on it.
Here’s the initial interactivity for the project:
I would map the physical distance the ultrasound sensor would measure (which I kept equal to the width of my computer screen) to the length of the canvas. Then, as the audience interact with the system by dragging some smooth object towards or away from the sensor, the swarm on the screen will accelerate in relation to the audience’s response. From a linear view, it should seem like the swarm in the background is following the object.
Part One: Creating the swarm
I coded up a Processing Class that, on instantiating, would yield an ellipse that accelerates towards a target. The ease of acceleration (how fast or how slow) was determined by multiplying the distance vector with a suitable scalar. The outcome was this:
A video posted by Apurba Akash (@apurbzz) on Jun 6, 2016 at 8:55am PDT
Part Two: The moment I realized my initial plan for interactivity wasn’t much productive
Once I got to tinkering with the ultrasound sensor, a bunch of things started going wrong. Firstly, setting the sensor to sense only the width of the computer screen was only slightly effective, at least my methods were I guess. Secondly, the sensor values fluctuated a lot, depending on what object is being placed in front of it. Irregularly shaped objects caused a lot of fluctuation.The whole interaction seemed unbalanced from a viewer’s perspective.
After some thinking, I decided to discard the whole “placing an object“ plan, and resorted to hand gestures - serious Jedi stuff yo! That way, any fluctuations in sensor value wouldn’t be too obvious to the viewer. Here’s how it went:
A video posted by Apurba Akash (@apurbzz) on Jun 6, 2016 at 1:01pm PDT
I was pretty satisfied with the results. Most people initially though this was Computer Vision. But, as the title suggests, I’m too broke to buy a Kinect, Also, since it took a while for the particles on the screen to accelerate towards the target, it felt as if I had to literally push the particles with my hand. It was interesting to see how the whole thing brings out certain sensations.
Part Three: Lesson Learned
Now, I plan to create a whole set of sketches where the audience can interact with the particles in different environments, such as inside a fluid, or in outer-space, etc. This just goes to show that experimentation always gives rise to new ideas, regardless of fruitful results.
