4C: Limited range for the head tracking (problems looking up)

[Michael Lehner]

Hello,

when using the head tracking function of the 4C, i can’t get close to the angle limits set in the ISE. When i disable head tracking and rely on gaze tracking alone, i can cover a much bigger up angle. (So, head tracking mode: none, gaze tracking mode: FreeView)

In contrast, i can very well look down when head tracking is enabled. I played around with the settings, but i seem to be unable to find a way to configure this the way i want it. I suspect that this is related to the devices position at the lower edge of the monitor. If this is true, it’s a design problem which can not be eliminated. If the problem is not in the devices position, could you please give me some pointers on what to set and adjust to work around these limitations?

[Mark Butterworth]

I have the same issue in elite dangerous. Head tracks down/left/right but will not track upwards.

I’m using extended gaze mode with dynamic head track mode.

[Michael Lehner]

Thanks for confirming that i am not alone with this. Now i am just curious if it can be solved. 🙂

[Grant [Tobii]]

Hi, We are aware of a problem with head tracking for some individuals when looking upwards. It depends a bit on the face how far upwards you can look without loosing head tracking. For some people it is barely noticeable, for some it can be more problematic.
 
This is due to the head tracking algorithms we are using. We are constantly trying to improve our products, so please be sure to check back for future firmware updates that will hopefully address this issue.

[Michael Lehner]

Yea, thanks for bringing this up, forgot to give feedback here: after upgrading to 2.7.0 i seem to have a bigger vertical range. It still seems to be a little smaller than what i can achive by going eyetracker-only, but as angling further up starts to ache my neck, i guess that is oki. Together with setting higher “speed” in the ISE i consider the range now to be good.

The only thing is: speed seems to affect the rate of how much the camera moves compared to how far i turn my head. Is this true of is this my imaginatin? If this effect is true, then the term “speed” could be replaced by something along the line of “ratio”.

[Michael Lehner]

Addendum: one tiny thing would still be nice to have: different view angle limits for up and down direction. But that’s not essential, it just would be nice if it’s possible to have that.

[Alex [Tobii]]

We are also working on an algorithm that would put more weight on gaze signal when head signal is reaching the limit.

[Author]

Posts