Todays opentrack works with edtracker and the joystick plugin to give you 3DOF TrackIR emulation. Spline based deadzone and nonlinearity settings, mappable tracker centering and on/off toggle.
The pic I posted earlier shows you the correct joystick axes for edtracker.
I have my current opentrack.ini available HERE
You have to set the EDtracker UI Tool to linear with 0 scale since all the deadzone and non linearity is done with spline curves.
You will probably want to change your hot keys. I currently use DEL for center and PageUp for on/off toggle. Both mapped to my X52. But I don't use the keyboard for controls in ED.
If it works adjust the mapping to whatever you want. It's quite sensitive without a filter but I didn't need one.
So we just need EDTracker's auto centering removed/adjusted?Todays opentrack works with edtracker and the joystick plugin to give you 3DOF TrackIR emulation. Spline based deadzone and nonlinearity settings, mappable tracker centering and on/off toggle.
The pic I posted earlier shows you the correct joystick axes for edtracker.
I have my current opentrack.ini available HERE
You have to set the EDtracker UI Tool to linear with 0 scale since all the deadzone and non linearity is done with spline curves.
You will probably want to change your hot keys. I currently use DEL for center and PageUp for on/off toggle. Both mapped to my X52. But I don't use the keyboard for controls in ED.
If it works adjust the mapping to whatever you want. It's quite sensitive without a filter but I didn't need one.
//self centering
// if we're looking ahead, give or take
// and not moving
// and pitch is levelish then start to count
if (outputMode != UI )
{
if (fabs(iX) < 3000.0 && fabs(iX - lX) < 5.0 && fabs(iY) < 800)
{
ticksInZone++;
dzX += iX;
}
else
{
ticksInZone = 0;
dzX = 0.0;
}
lX = iX;
// if we stayed looking ahead-ish long enough then adjust yaw offset
if (ticksInZone >= 10)
{
// NB this currently causes a small but visible jump in the
// view. Useful for debugging!
dzX = dzX * 0.1;
cx += dzX * 0.1;
ticksInZone = 0;
dzX = 0.0;
}
}edtracker2:185: error: 'TrackState_t' does not name a typePhew, ok, just caught up on several pages of action after being away on hols for 4 days, so sorry that there's been no official response on here from the EDTracker bunch... I think I am the unofficial spokesman
So, to clarify...
I *was* looking into the Opentrack issues but I have to pass on massive thanks to Antmax who, in my absence, has done it all for me, liasing with sthalik (the developer of opentrack) to figure out what the issue was and fix it.
It's clear that the EDTracker code doesn't need to auto-centring aspect when in use with Opentrack, so Rob is now aware and is working on making it a config option in the EDTracker GUI so you can flick it on or off. I think I might also look at changing the auto-centring so that it applies over a longer period of time than it currently does, as it is obviously causing some people issues. Magnetometer integration is something I am looking at too, which fits in the same sort of code area, so I can probably kill two birds with one stone there.
As a short-term fix, I can build a non-centring version of the firmware and put it up on the download site so you can flash it in with the GUI tool and command line tool, for those who are eager to try it out with Opentrack.
Great work guys...
Cheers,
Dan
Great work guys...
Confused - Isn't the setting of the sensitivity and the future auto-centering yes/no option autonomous from the drift testing/compensation?
I think what's needed is the auto-centering feature is just toned down, so that it's not so immediate, but averaged out over a longer period of time. Easy enough a change
I think your idea of simply testing over longer periods could improve it.
