CAD Design

CAD Design finished!  Incorporating the origin-centric design, having the camera rotate around a fixed point instead of a neck point.

Parts were designed in SolidWorks 2015 with 3D printing in mind, keeping rounded parts in the flat plane as much as possible to reduce resolution loss when printing with a larger step (0.4mm).

Mechanics of Head Movement

So I'm continuing my research and planning prior to any design and purchases, and one of the more interesting issues I've come across is methods of simulating head movement.

Initially, my plan was to create a simple gimbal that would rotate the dual-cameras around the "origin", keeping them relatively fixed.  As I continued with this train of thought, I realized that this is not how our eyes move relative to our heads. Our head moves relative to our neck, and our eyes are at a fixed distance from the "center" of our heads and a fixed height from our necks. If you look around now, you'll realize that our "origin of view" is not fixed as we sometimes imagine it to be. It moves in an arc at a fixed radius.

The real question to consider here is how much this difference in movement will affect comfort when using the camera.  Abnormal movement might contribute to an "out of body" feeling, if our origin of view does not change as we turn our heads like we are used to.  Depending on time and funds available, I hope to investigate both.

After some brief thought, I have come up with these three options....

Camera Rig Sketches, P - Pitch servo, R - Roll servo, Y - Yaw servo

Option 1 was an original prototype for the "natural view" solution, rotation around the base, pitching from the center, and rolling at the cameras.

Option 2 gives the most realistic interpretation of the rotation about the neck, where the origin of pitch is not located below the eyes (close to where the neck should be).  The angle of pitch, measured between the origin and point of cameras, is not in line with the direction of the cameras, which is more representative of how our eyes are oriented.

Option 3 is my original design, a simple horse shoe support to prevent vibration, wobble, or overshoot which would lead to discomfort.  In this setup, the cameras revolve around the center, marked by the "R" or "Roll" servo.

Will update with more details to come, wanted to get this down as I thought about it!

New School Year

With the start of the new school year, I'm looking to build this project into my schedule and make some progress on design and construction.  This week  I've spent time familiarizing myself with properties and limitations of servo motors before I make any purchases, looking to define my needs and objectives and how these will fit in with them.

I've also been looking into ways of interfacing the Rift and servos/motors/etc, current winning plan is using an Arduino microcontroller to run the servos, with Unity (and a special software pacakge) as a go-between to interface the Rift and Arduino.  Unity supports the Rift and Arduino with the correct packages, so this seems like a logical way to reduce the complexity in software.

Starting my Robotics course at UC this week! Looking forward to see how it connects to this project and seeing what I can leverage to help me along my way.

More updates to come, exciting time to be in VR!

- Smith

Welcome to VP Camera!

First post!

This blog is dedicated to the research and development of a virtual presence camera, using an Oculus Rift and various components to create a webcam capable of 3DOF motion, integration with the Rift tracking, and stereo view.  I'm definitely an amateur, so this is going to be a learning experience, hopefully I'll have something to show by the end!

I hope to make semiweekly posts tracking my progress, posting photos, etc.

Each post I hope to address these points:
a) progress made
b) what worked
c) what didn’t work
d) what the next step will be.

Stay tuned for some amateur mistakes and some really cool tech!

- Smith