I would like to figure out if the MT360 still have limitations for which angles the viewer is looking from.
In the Generator application, we see a green circle, and a red pitch indicator, but no roll indicator (we did see that in the Early Access version). It also seems to us that the object recognition works mostly (maybe even only) when we hold the device at a certain roll angle corresponding with the up-vector the virtual model had when trained.
My question: Does MT360 have limitations on the possible yaw/pitch/roll angles at which objects can be recognized? Especially roll seems an issue.
And if so, is there anything I can do to increase my chances of recognizing an object? Ideally, we would like to recognize objects from any angle, including roll.