FreeTrack Forum

Hello all!

Great piece of software. Congratulations!

I tested it briefly with small pen LED light strapped to my headset, and Lock on seemed totaly new experience.


Just a quick question. FAQ states this:

Which point model configuration should I use?

There isn't a great deal of difference between them, the 3 point clip is a vertical configuration for headphone mounting and the 3 point cap is for baseball caps. The 3 point clip has some extra calibration controls to minimize unwanted translation due to the off-centre position of the model.

Cap model LEDs move much more with head movement than clip model. Camera resolution is just 320 x 240, so small movements on clip model could be lost (smaller than 1 pixel on camera). Has anyone tested the accuracy of both shapes? Is the size of assembly important? I have 24" monitor almost meter away from me, and camera STX has quite a wide field of view - which I like, because my head stays in camera view even if I move around a lot.


And completely unrelated question:

I don't know if I ordered correct resistor. My tracking system assembly will be built onto headset. I have Logitech Communicate STX Webcam, I have already removed the IR filter (camera still works fine for camera use if I remove the visible light filter). I ordered the IR LEDs (SFH485, without P, but I will file the ends flat), and resistor, and I will hook this to 9V battery.

I used the LED Assembly Wizard, and it calculated 240 Ohm resistor for serial connection - LEDs will have 1.5 V and 20 mA current. I entered decimal value with period (1.5). But if I enter the value with comma (1,5), I get 300 Ohms - I think the page uses period, not comma, because I also get 300 Ohms if I enter 1 V.

I hope 20 mA current will be enough for SFH485, they are rated for max 100 mA and 1.5V, same as SFH485P. My camera has IR filter removed and I will use developed film to block the visible light (which is already very low at minimal exposure and 30 fps).

Thank you for the answers!

Bwaze @ 02/02/2008 - 19h33 a dit:

Cap model LEDs move much more with head movement than clip model.

To measure rotation, an observable change in the relative distances between the points is required, it doesn't matter how much, or how little the model translates during the rotation. The rotation accuracy is dependent on the depth of the model and in this respect the clip and cap are the same.

Is the size of assembly important?

Only the size of the model in the camera frame.

Thanks for the quick reply!

I think I will make both setups before I choose the permanent one, just quickly out of cardboard and test them a bit. I'll post some pictures and my findings.

I hope all the parts will arrive tomorrow, I can hardly wait.