As we all know there are only three ways one can move a camera. They can be panned, tilted and rolled. Therefore it would seem logical that 3 axes would be all that would be needed to be incorporated into a gimbal design. This would be so if not for the annoying mechanical constraints that must be addressed.
The traditional approach to high-level gimbal design is to have a second redundant axis for azimuth (pan, yaw), tilt (pitch) and roll with 3 coarse outer axes and 3 direct drive inner axes. This results in a 6 axis system. What it does not avoid though is a condition known as “gimbal-lock” which briefly is a relationship between the base mounting and the camera or sensor direction in which the camera loses the ability to be controlled or stabilized in a certain direction. (A system with this problem can still “look straight down” contrary to the understanding of some). In order to truly solve this “gimbal-lock” challenge, an additional outer primary axis is required.
As to the number of axes, the redundant double axis approach has traditionally been used to add an additional layer of stabilization. Typically, the coarse outer axes have employed gear or belt drives to steer out of the way of the inner “fine” axes that employ a direct drive system.
In Oculus we have investigated obtaining equal or superior levels of stability in a single direct drive stage for each of its 4 axes. We have achieved this and in so doing have created a much lighter and more responsive system that can look anywhere regardless of mounting position. This approach also opens up aesthetic possibilities that simply have not been contemplated before…. but…. they will be very soon.
…more to come.