I enjoy how new hardware paradigms are being pioneered. Early GUI concepts such as the Teleporter widget (Xerox PERC Alternate Reality Kit, 1985) inspire me to rethink the common approaches to interface design.

I've screened some references, to check how different graphics look on the glass, and then started sketching. 

I was exploring a wide variety of visual systems, ranging from old computer shells, to Soviet aircraft HUDs. During my graphic archaeology practice, every item I found one that was special to me, I would reflect their aesthetics in my design.

As soon as I had enough ideas for creating visual vocabulary and layouts, I moved to production.

I wanted my interface to start with a hub menu, used for accessing five different menu items. Each menu item is a sequence demonstrating different UI concepts, and featuring different gestures. 

After drafting the sequences I made them into five still-o-matics, and started iterating on timings, and on hand choreography. I output the still-o-matics onto the glass to rehearse the gestures.

I'd been filming myself to review if my hands were moving nicely, and weren't obstructing the view. After review, I made all the necessary adjustments to the animation blocking.

Next goes the visual vocabulary. I have created plenty of assets: shape libraries, widgets, boot sequences, loading screens, icons, preloaders, buttons, gauges, grids, flowcharts, scrollbars, etc.

All of the above laid on the foundation of five layouts. Each of the five sequences consisted of four phases, smoothly transitioning from one to another.

I was going to film each of these sequences separately, and turn this project into a five episode miniseries.

However, creating assets and putting them together took me a while! And since I haven't even started animating, filming and editing, I made a decision to leave only one sequence out of five, and focus on polishing it 💀

As soon as I had my design looking cool and unusual enough, I proceeded with setting it in motion. I had a lot of fun making things flow, and seasoning them with mecha anime gubbins, and Sega Saturn video display processor artefacts.

Here is the complete sequence of Episode V:

The end product is a video, so I decided not to spend my time engineering an input system, and​​​​ focus on creating graphics. Simulated touch input is a technique I use to cut R&D costs, and to enjoy playing with different ideas, seeing them work.

However, since the sequence became quite long, I considered creating an interactive playback system to help me with filming.

I used a Leap Motion sensor for touch registration, and cueing the clips. Unfortunately, hand detection wasn't precise enough in my setup, so I dropped this idea.

I started with filming several tests, figuring out the best lighting and camera setup for the given location, and my equipment. 

I was experimenting with different types of camera movement and shot composition. Furthermore, I also did a few takes in slo-mo, in an attempt to combat DLP projector artefacts, and to give myself some headroom for gestures performance.

After numerous takes, I did a few edits of my footage, trying to tell different stories. Then I changed my mind for a continuous shot, helping me to focus on what's important. 

And that's how it was done