My opentrack settings

April 13, 2022

April 13, 2022

I started out trying to tweak the response curves, but what really improved things for me was to actually calibrate the setup in Opentrack. The function is rather well hidden behind an additional layer of tabs and I was wondering why people kept talking about calibration when there was no button to start it. ?

I also played around with the smoothing- and deadzone-sliders, increasing both by quite a few clicks for roll, pitch and yaw, while leaving X,Y and Z pretty close to their default settings. And finally I upped the camera's FPS from 60 to 75, but I'm not sure how much impact that has (probably negligible).

Before I had difficulty keeping my view steady around the center (watching video-clips of me flying with those older settings gives me nausea, it was that twitchy ). Now it's still not ideal and needs a little more tweaking (center view still isn't quite stable enough to reliably look down the gunsight and use it without switching to the fixed gunsight view), but overall things are a lot better and way more stable now.

Plus I can finally look behind me with consistency and reliably move my virtual head to get a view past the, say, 190's headrest. First thing to achieve that was to mess around with relative translation but also to calibrate my setup. Before calibration, I think that the space observed by the camera was way smaller than it should've been, thus small inputs on my part resulted in overly harsh/quick movements of my virtual head and the camera "lost" the LEDs on my Delanclip way earlier than it technically should.

In this current state, the performance is nearly as good as it was with TrackIR5/Pro Clip on my old system, and I'm pretty sure I can get it up to that level with a little more tweaking.

S.

Edited April 13, 2022 by 1Sascha

April 16, 2022

Try Kalman filter its a lot faster

Also I turn off head roll

And you can see I have dead zones around the scope

Only problem I'm really having at the moment is when I'm looking straight up it feels janky tracking planes as they move from left to right
but idk what to do about that

April 16, 2022

4 hours ago, brahguevara said:

Try Kalman filter its a lot faster

Just tried it, but it felt too twitchy for my taste.

I'm still trying to find settings that will create sort of a "suction zone" around the center position. So that my view will rest on/near the gunsight and take more head input to start moving away from the center.

BTW: What exactly does that "neck displacement"-thing do that you have active in your settings?

April 16, 2022

5 hours ago, 1Sascha said:

Just tried it, but it felt too twitchy for my taste.

I'm still trying to find settings that will create sort of a "suction zone" around the center position. So that my view will rest on/near the gunsight and take more head input to start moving away from the center.

BTW: What exactly does that "neck displacement"-thing do that you have active in your settings?

Here is how I have my yaw and pitch response curves set. This works for use with face tracking. Something similar would probably work with IR.

April 21, 2022

On 4/17/2022 at 12:36 AM, 1Sascha said:

Just tried it, but it felt too twitchy for my taste.

I'm still trying to find settings that will create sort of a "suction zone" around the center position. So that my view will rest on/near the gunsight and take more head input to start moving away from the center.

BTW: What exactly does that "neck displacement"-thing do that you have active in your settings?

see the flat line 0-5
that means I can move my head 5cm either side of center and the camera will still be in the middle of the gunsight

April 21, 2022

39 minutes ago, brahguevara said:

that means I can move my head 5cm either side of center and the camera will still be in the middle of the gunsight

I don't think we're talking about the same thing.

I did in the meantime find out how to set additional points in the response curves. Played around with those for yaw and pitch, but I found that with a "suction zone" in the center, the in-game response feels even more awkward and unnatural to me than the twitchy-ness I get with straight, more linear response curves.

*shrug*

I was talking about this setting here:

Seems to me that activating it turns OT-behaviour more into what I was used to from my TrackIR5. Makes it feel a bit more "natural" I suppose, since, when I activate it, my virtual eyes aren't in the same place as my virtual neck anymore.

Still fiddling with the distance setting in there, but for now, 5cm seems to work pretty well.

S.

Edited April 21, 2022 by 1Sascha

April 22, 2022

18 hours ago, 1Sascha said:

I don't think we're talking about the same thing.

I did in the meantime find out how to set additional points in the response curves. Played around with those for yaw and pitch, but I found that with a "suction zone" in the center, the in-game response feels even more awkward and unnatural to me than the twitchy-ness I get with straight, more linear response curves.

*shrug*

I was talking about this setting here:

Seems to me that activating it turns OT-behaviour more into what I was used to from my TrackIR5. Makes it feel a bit more "natural" I suppose, since, when I activate it, my virtual eyes aren't in the same place as my virtual neck anymore.

Still fiddling with the distance setting in there, but for now, 5cm seems to work pretty well.

S.

O i get ya

Yea I only turned that on recently seems to make it feel a little bit better

I use grassmonkey simulations puck
so the 3 leds end up at my eye level

but the 5cm does feel better might play around with bigger values or even negative values at some point

also try adjusting these values it makes things work better
especially the bottom two

Edited April 22, 2022 by brahguevara

April 22, 2022

5 hours ago, brahguevara said:

or even negative values at some point

I'm probably overthinking this at this point, but that seems like a weird idea ... if it really works to emulate real life. Where your eyes are always "in front" of your point of pivot (your neck). If "zero cm" means you're on the same plain as your neck, using negative values sounds to me like, in real life, you'd have to push your eyes out of the back of your skull and mount them on a rail back there or something. ?

Of course: It still could have a positive effect in a simulation environment, but thinking about what it would mean in real life is kinda bizarre... and a little scary... ?

S.

November 23, 2022

Updating this with adjustments I have made.
Also grassmonkey simulations now makes the best IR head tracker. Using a custom camera with much wider FOV and speed 240 hz compared with the competition.
I have owned the ps3 camera version and puck for over a year now and are very happy with it.