For the final presentation, I needed darkness to show the hologram optimally. But I realized that wouldn't be possible. So I made view boxes (complete with googleyes of course) so that viewers could put place the pyramid on their smart device, place on top the view box, pop out one of the google eyes, and view the hologram in complete darkness.

I first got all of the dimensions of the smart devices I wanted to accommodate. I then made a Rhino file for lasercutting the view boxes, where blue lines represent score marks and red represent cuts.

The first thing I wanted to do was create the transparent electrode which would transfer the 40,000V from the ignition coil to the object, but allow for the view to see it due to it's transparency. So I got a Indium Tin Oxide (ITO is a clear conductor) coated glass plate and began making the transparent electrode.

For the electronics inside the box, I wanted to clean up some of the boards of the previous weeks to incorporate any previous hacks, cold connections, loose wires, and unnecessary components. The most significant change I made was with the audioboard. Originally, I used the audioboard with an electret microphone which picked up a sound input from a speaker. But since I was limited by the gain of the opamp I was using, that I meant I would have to blast the speakers - which would be inappropriate to do for the final presentations. So, I decided to use the analog DC stereo output of my headphone speakers instead, which was about 10x higher in voltage and didn't oversaturated my mosfet. And - of course, I had to add the logo to the mill file!

Of course, I needed an enclosuse that could both safely house all of my high voltage electronics, as well as keep the object in pitch darkness to amplify the kirlian electrophotography effect. Using rough dimensions from previous protoyps in prior weeks, I saw that a 9'' by 9'' box worked well in the XY plane and the webcam seemed to work best (in terms of autofocus) at about 6 inches of height.

As always, branding is a key component of a successful product design. So, I wanted to make a simple brand identity based off the name I came up with, the "ShockBox"...Since yellow and black together often connote danger, I thought that would be an appropriate color scheme for a 40,000V project! I also used thunderbolts in a box formation as the standalone logo and motif throughout the sides of the box.

After laying out the design along the whole box, I prepared the file for the vinyl cutter in grayscale. I made sure the artboard width matched the width of the vinyl roll, an to go through a test runs to optimize the force on speed of the cutter (in the end I used 80gf and 1.5cm/s).

Here is a copy of the final tornado based server code I wrote up. The first thing I do is import all of the packages needed for python. I then ask the user to input two values: (1) the brightness adjustment and (2) the size of the object so that the webcam can crop down to the object. I then initiate three classes, a long poll class, an image handler class, and a main handler class - each with their own defining functions. The long poll class manages the server side, waits for an image, and prints out how many clients it's serving. The main handler class is in charge of rendering the html file, and finally the image handler gets a new image from a long-poll object and updates it to the jpg header if the stream is closed.

This section is the final image processing section of the server code. Here I initialize a camera capture object with opencv videocapture, I then run three functions on it: (1) triptastic (2) rotated and (3) rotate3D...Triptastic simply takes the webcam video and makes it into four reflected images, while rotated then lays out those images in a 90 degree rotation transformation from one another. Finally, the rotate3D function is the pseudo-animation which rotates the object in space with an optional skew to make it distort at specific angles.

The last thing I needed to do on the image processing side was initiate the webcam, start up the app, and write the get image function for receiving the image from the camera object.

For the sound processing, I made a quick python "juke box" code as I call it that allows a user to select a music file (.wav format is best) from a "sound files" directory on the desktop.