FreeTrack Forum

First,

Thanks for taking the time to create a cool product.

System specs:

Vista 64
6Gb Ram
Intel i7
Logitech QuckCam 9000 pro, unmodded.

I have the software installed, and it seems to be working. However, I am seeing minimal movement of the 3d skull, when I physically move my head.

I tried increasing sensitivity, but that didn't seem to help very much. Movement still feels pretty restricted. I can see the numbers under the skull changing, so I know the software is processing data.

Any thoughts on the following questions would be apprieciated:

- How best to be certain the model dimensions in the software, correspond to the physical measurements required.

- Whether it is possible to increase the 's' factor, shown below the skull.

Any other general thoughts or hints.

Thanks for taking the time to read, and reply to this message.

Does this help anything?
http://forum.free-track.net/index.php?showtopic=2223

Post screen-shot, write what model of LED configuration you use.
You could adjust model size under 'Model' tab if that was the issue.

Thanks for pointing me to that thread.

I did try a single-point test, with a remote I had around.

I saw an immediate difference. A lot larger numbers, and very obvious movement of the head in the frame.

It is definitely true that the further I pull the light from the camera, the further I have to physically move it, to get the desired displacement, but I was still able to get the head to the extreme left/right/top/bottom position. Again, this is with the sensitivity sliders all the way to the right.

My camera resolution is set to 320 x 240, I420 compression, 30 fps.
Room was dark, except for the monitor and the remote's LED.

I tried 640 x 400, and didn't see any appreciable difference.

However, distance from the camera is a definite factor. Closer in, I don't have to physically move the remote as far either direction to get the endpoints. Further back, I have to move it about twice as far to get the same result.

So, I'm still curious whether there is a suggested maximum distance you should have beetween your sensors, and the camera.

For what it's worth, I typically sit about two and 1/2 keyboard widths from the physical monitor, as my chair will not fit under the desk, and the desk  is just wide enough to have room to keep an old VCR and the keyboard on top.

That camera is expensive one. I'm not familiar with it.
You can check if it have zoom ring or other way controlling field of view. Maybe picture size is just too big. In that case LED needs to travel greater distance for small movement.

I'm about 80cm from camera. It work flawlessly. You are about same distance also.
At 80cm from my cam, picture frame is 50 x 38cm (I took piece of rope and stretch it across with my hands. Later, I had measured how long rope between my hands was)

Also, you can check if you messed something with 'Global profile'.

Well, sorry about my English

Hi,
I'm trying to set up FreeTrack for the first time, and I'm having exactly the same problem.

I've built a 3 Point Clip model and I'm using a wiimote as IR camera.

Well, the problem is just as described on the first post. I rotate the model in the yaw axis as much as 90º but on Freetrack it shows movements as small as 1º or 2º.
I have set the curve for Yaw to 1:1 to see what angle FreeTrack is calculating from the camera points and is like that, 1º to 2º, whereas the model is almost 90º.

I've selected the correct model and put the dimmensions of the built model in milimeters. But it's like if FreeTrack thinks that the model is a lot bigger that it really is. I've checked and rechecked dimmensions a thousand times.

Played a lot with the curves before realizing what is the problem.

I've seen that this thread ended without a solution. I hope someone can help anyway.

I also have a 3 point CAP model that I use to track my VR goggle with my software. I haven't do a test with it but I'll do it as soon as possible and will post if there is any difference.

Ok, the same problem with the 3 point cap model.