I have found that where you position the TrackIR will have an effect on the "translation" axises (X, Y, and Z).
I was planing on placing the device on a shelf above my monitor to give a bit more distance and allow more movement. It turns out this causes leaning forward (Z) to be read as moving down (Y).
It would be nice if there was a setting where you could adjust this. Like a slider that you can set inches above/below eye level and distance left/right of monitor.
Translation Offset
Re: Translation Offset
It shouldn't make a difference, as long as the reflectors are ...
1 - Centered on the Tracking screen
2 - Also have the software tracking re-centered.
1 - Centered on the Tracking screen
2 - Also have the software tracking re-centered.
Re: Translation Offset
I have noticed this also.
In dogfighting games, when I lean forwards to zoom in on the sights, because the TIR unit is at the top of my monitor which is above the top of my head or even worse with the trackclip, lower down on the side, then it also moves the view down a bit and puts my sights out.
I too would like to be able to do as the OP says and move the TIR unit up and back, and tilt it down - to give it a wider field of view, and allow me to get closer to the monitor without it losing lock and going nuts.
Maybe some calibrastion system would be in order? Like a joystick calibration - move left, move up etc. That way, you could probably even have the unit any way up (eg attached to the left of the monitor, on it's side) and at any angle and it would still know which way was which.
In dogfighting games, when I lean forwards to zoom in on the sights, because the TIR unit is at the top of my monitor which is above the top of my head or even worse with the trackclip, lower down on the side, then it also moves the view down a bit and puts my sights out.
I too would like to be able to do as the OP says and move the TIR unit up and back, and tilt it down - to give it a wider field of view, and allow me to get closer to the monitor without it losing lock and going nuts.
Maybe some calibrastion system would be in order? Like a joystick calibration - move left, move up etc. That way, you could probably even have the unit any way up (eg attached to the left of the monitor, on it's side) and at any angle and it would still know which way was which.
Re: Translation Offset
You already have the option to moint it to the side, upside down, to your left, right, etc..
If you can go into the TrackIR software View->Tracking->Settings. You can then set the rotation here; 0, 90, 180, and 270 degrees.
If you can go into the TrackIR software View->Tracking->Settings. You can then set the rotation here; 0, 90, 180, and 270 degrees.
Re: Translation Offset
I don't think either of us are wanting to change the orientation of the actual TrackIR device. My TrackIR is rightside up, but not directly infront of me. Instead it is about a foot higher.
I no longer believe a new setting is nessesary, but rather a better algorithm. I think I can describe the problem so you can reproduce it. For this test I used my TrackClipPro, but the vector clip should give similar results. Here are the steps:
1. First open the TrackIR software and make sure the tracking, guages, and heads windows are all open.
2. Hold the TrackClip directly infront of the TrackIR sensor, and face the TrackClip to the TrackIR and center the software.
3. Slowly move the TrackClip directly toward the TrackIR and observe the Y and Z guages.
[image]http://img.photobucket.com/albums/v59/C ... slide1.jpg[/image]
4. Now, return the TrackClip to the original position further away from the TrackIR. Angle the TrackClip downward to simulate the TrackIR being higher than eye level. Be sure the LED lights don't overlap in the tracking window. Recenter the TrackIR software.
5. Again, slowly move the TrackClip DIRECTLY toward the TrackIR sensor and observe the Y and Z guages.
[image]http://img.photobucket.com/albums/v59/C ... slide2.jpg[/image]
RESULTS: When I did this test I found in both instances Y axis movement would be minimal and Z axis was the only noticable movement. This means that only the zoom "Z" axis what effected. Why is this bad?
Well, imagine if in Step 4 we had actually set the TrackIR high enough that it would see the TrackClip at the same downward angle if the Trackclip was facing the center monitor. You would have to move your head up and forward to only effect the zoom "Z" axis.
Now try this:
Return to step 4, but at step 5 instead of moving the TrackClip directly toward the TrackIR, move it forward at the angle that it is facing. Move it downward at the angle it's facing and watch the Y and Z axes again.
I usually observe about the same Y movement as Z movement!
This test is to observe the differences when the TrackIR is set higher than eye level. To test lower than eye level, in step 4 angle the TrackClip upward. I believe this problem can also occur if the TrackIR is places farther left/right of center view too, but I don't care to test it.
Conclusion: The TrackIR algorithms are interpreting translational (X, Y, and Z) movement relative to the TrackIR sensor instead of relative to the TrackClip. They don't seem to take into account that the TrackClip may not be directly facing and at the same level as the TrackIR sensor. My original post was to correct this manually in the software, but I think it could be don't automatically since the software knows which direction the TrackClip is facing.
I hope I've made myself clear, because this was a somewhat difficult thing to describe. I hope you can replicate it so you can show the dev team.
I no longer believe a new setting is nessesary, but rather a better algorithm. I think I can describe the problem so you can reproduce it. For this test I used my TrackClipPro, but the vector clip should give similar results. Here are the steps:
1. First open the TrackIR software and make sure the tracking, guages, and heads windows are all open.
2. Hold the TrackClip directly infront of the TrackIR sensor, and face the TrackClip to the TrackIR and center the software.
3. Slowly move the TrackClip directly toward the TrackIR and observe the Y and Z guages.
[image]http://img.photobucket.com/albums/v59/C ... slide1.jpg[/image]
4. Now, return the TrackClip to the original position further away from the TrackIR. Angle the TrackClip downward to simulate the TrackIR being higher than eye level. Be sure the LED lights don't overlap in the tracking window. Recenter the TrackIR software.
5. Again, slowly move the TrackClip DIRECTLY toward the TrackIR sensor and observe the Y and Z guages.
[image]http://img.photobucket.com/albums/v59/C ... slide2.jpg[/image]
RESULTS: When I did this test I found in both instances Y axis movement would be minimal and Z axis was the only noticable movement. This means that only the zoom "Z" axis what effected. Why is this bad?
Well, imagine if in Step 4 we had actually set the TrackIR high enough that it would see the TrackClip at the same downward angle if the Trackclip was facing the center monitor. You would have to move your head up and forward to only effect the zoom "Z" axis.
Now try this:
Return to step 4, but at step 5 instead of moving the TrackClip directly toward the TrackIR, move it forward at the angle that it is facing. Move it downward at the angle it's facing and watch the Y and Z axes again.
I usually observe about the same Y movement as Z movement!
This test is to observe the differences when the TrackIR is set higher than eye level. To test lower than eye level, in step 4 angle the TrackClip upward. I believe this problem can also occur if the TrackIR is places farther left/right of center view too, but I don't care to test it.
Conclusion: The TrackIR algorithms are interpreting translational (X, Y, and Z) movement relative to the TrackIR sensor instead of relative to the TrackClip. They don't seem to take into account that the TrackClip may not be directly facing and at the same level as the TrackIR sensor. My original post was to correct this manually in the software, but I think it could be don't automatically since the software knows which direction the TrackClip is facing.
I hope I've made myself clear, because this was a somewhat difficult thing to describe. I hope you can replicate it so you can show the dev team.
Re: Translation Offset
added graphics
Excuse the mspaint TrackIR as I couldn't find an image of it from the side.
Excuse the mspaint TrackIR as I couldn't find an image of it from the side.
Re: Translation Offset
Noted....