Strange axes dependecies above 4.0.18

[steku]

I'm still using 4.0.18 software with my TrackIR3 Standard and VE (clip mounted to TrackHat). This is because when I try any of new versions I'm getting strong dependencies (in FS9) between axes. My TrackIR3 is mounted slightly above my head position (5 cm) at distance ca. 70cm.

E.g.:
When I (sitting carefully still) rotate my head up/down in any version newer than 4.0.18 my point if view is moving forward (when tilting my head down). I tried to edit new sensivity (less sensivity in forward - backward motion) curves from scratch but to no avail (after all sometimes I need this linear motion, when I move my head forward-backward and I can't lower sensivity too much). This interaxis dependency makes my TrackIR useles - for example I can't look down at my trim tab setting bacause I'm instantly moved closer (too close) to the panel.

1. Could you suggest me what's wrong with post-4.0.18 versions in my case?

2. Could you post some information what's new in 4.1.028? I don't want to install this new version, bacause I want to protect my precious 4.0.18 configuration, but I have to consider it after reading about new features.

[von Baur]

I don't think there's anything wrong, steku. I experienced the same thing and realized that, because the pivot point is your neck, as you tilt your head down your forehead (where the reflectors sit) is moving forward in three-dimensional space. Likewise as you tilt up it moves back. My solution wasn't sensitivity, but creating a dead zone in the zoom. That way I can tilt my head without having Track IR read it as moving in or out.

[steku]

I'm sure Natural Point have taken into account the human anatomy and position of clip when placed on hat and all this makes a foundation for equations contained in TrakIR software. This is especially true now when new version of software is available with two different clip positions - on hat and on headset with two different sets of equations.

But my question is: Why I see more interaxes dependencies in any TrackIR version after 4.0.18? What was changed after 4.0.18?

[rodom]

A lot has changed with the change.log since about 9 months ago when 18 came out. I do know that things are more sensitive with the profiles and with that being said you might want to make more deadzones or edit the dead zone setting itself to suit your needs.

[rodom]

Do you know which version this started happening? Have you tried all revisions in between the current and the version 4.0.18?

[Barred]

quote:Originally posted by NaturalPoint - Rick:
A lot has changed with the change.log since about 9 months ago when 18 came out. I do know that things are more sensitive with the profiles and with that being said you might want to make more deadzones or edit the dead zone setting itself to suit your needs.The easiest way to desensitise the profile and make/increase the dead zone is to enter the editor for your profile and click on 'Edit Yaw'. In the top right there is a up/down arrow for 'edit entire curve' bring the whole curve down antill you have a small flat spot in the centre and then if needed you can move the curve back up a notch or two. Then copy to all axises in the bottome left.

If you have left the curve lower than before then you have given yourself a smooth dead zone and a slower profile at the same time.

If you have not raised the profile after the dead zone and it is still too fast then you will need to edit the points from the centre outwards.

[steku]

Thanks for the advice. I think I'm pretty good at defining the sensivity curves as a long time TrackIR user.

After 4.0.18 I tried versions 4.0.20 and 4.0.21. The problem with stonger dependency between pitch and Y and Z axes was introduced in one of mentioned versions (4.0.20 I think). But I didn't test 4.0.19.