Neither does Ed Tracker, and you don't need a camera to use it.
Aruco doesn't need an accelerometer, magnetometer, or a gyroscope.
The "device" itself needn't be connected to a power source or anything else in particular.
diy track ir with webcam?anyone?
- Thread starter JamesBondage
- Start date
Aruco doesn't need an accelerometer, magnetometer, or a gyroscope.
The "device" itself needn't be connected to a power source or anything else in particular.
At the moment I'm testing FacetrackNoIR with a Logitech c615 webcam. Funny thing is, it works really good in Warthunder, does not work at all in X Plane 10, and it's kind of stuttering (sometimes) badly in E: D -.-
Need to fiddle with the curve-settings, I think. Atm I'm elite'ing without tracking software. But my time will come *evil laugh*
Need to fiddle with the curve-settings, I think. Atm I'm elite'ing without tracking software. But my time will come *evil laugh*
It is fun and immersive but I've come to conclusion that once your sufficiently good at interpreting the scanner and target info panel its of little or no practical value in combat, in fact if anything it can be a disadvantage because most of your instrumentation is out of view when your tracking.
I just built a 3-point clip for use with Freetrack and it works great. Using modified PS3Eye with CLEye driver. Made the clip out of 6 identical precut pieces of cardboard glued together like plywood and press-dried overnight. Electronic components from DigiKey.com. The battery box and cam & stand from Amazon. Total cost for all hardware was about $40 including shipping.
Here's a video of the first test flight with it, helpful links in video description: http://youtu.be/1fEnAzDS3hs
Circuit
Pre-solder stage
Soldered and working
Attached to headset with binder clip - just to test. I will wrap in black tape once I am happy with it.
PS3Eye, IR filter removed and IR band-pass filter (a piece of floppy disk) attached
Can't wait to try it out in combat when I get a minute!
Here's a video of the first test flight with it, helpful links in video description: http://youtu.be/1fEnAzDS3hs
Circuit
Pre-solder stage
Soldered and working
Attached to headset with binder clip - just to test. I will wrap in black tape once I am happy with it.
PS3Eye, IR filter removed and IR band-pass filter (a piece of floppy disk) attached
Can't wait to try it out in combat when I get a minute!
Last edited:
I've tried them all so far with my PS3 cam and IR lights. They were nice in that you got full 6DOF but having to maintain a fixed position in front of the camera can cause a lot of aches over time and on a older gaming system use a lot of resources tracking video at 120fps. Plus there is always a balancing act trying to get no stutter and lag. The hit on my system with a GTX650 was about 8FPS in most games.
Edtracker originally had the limitation of only 2DOF with native ED control settings which only allows for 2 axis rotation like mouse look. This week opentrack had the joystick input tracker fixed and Edtracker works in 3DOF rotation. Doesn't need filters and uses hardly any resources and doesn't require your head in camera all the time, so posture and position isn't a big deal.
It also uses almost 0 computer resources and is more precise than TrackIR. If you don't need the x,y,z head movement and can live with just head pitch, yaw and roll EDtracker is the way to go. No desk clutter with cameras and funny IR clips.
I tried Aruco too but it had trouble tracking at some angles and fixing the image to my head was a bit awkward.
Edtracker works very well, but for the time being you have to disable the auto recenter in the EDtracker arduino source code for it to work properly. That should be fixed in the next week or so though.
Edtracker originally had the limitation of only 2DOF with native ED control settings which only allows for 2 axis rotation like mouse look. This week opentrack had the joystick input tracker fixed and Edtracker works in 3DOF rotation. Doesn't need filters and uses hardly any resources and doesn't require your head in camera all the time, so posture and position isn't a big deal.
It also uses almost 0 computer resources and is more precise than TrackIR. If you don't need the x,y,z head movement and can live with just head pitch, yaw and roll EDtracker is the way to go. No desk clutter with cameras and funny IR clips.
I tried Aruco too but it had trouble tracking at some angles and fixing the image to my head was a bit awkward.
Edtracker works very well, but for the time being you have to disable the auto recenter in the EDtracker arduino source code for it to work properly. That should be fixed in the next week or so though.
I created a little Construction and Setup Guide for anyone wanting to learn more about building your own Freetrack IR Headtracker
Several points -- I'm using paper marker tracking since spending PLN on clips and stuff just seem wrong here.
Marker requires a bit more res, and has no builtin light. 640x480 a must. Sun dies quickly, so switched 640x480 to @ 30 Hz temporarily. It worked without issues, that is, no perceivable slowness.
> Plus there is always a balancing act trying to get no stutter and lag.
Accela rev. 5 user-facing options make more sense than rev. 3-4 and even 2, it's all dbaarda and his explanation of regression+moving averages+what he did with EWMA. Accela mk5's now much closer to what was 2012' DZ1 filter from which accela mk2 came. Here's some stuffies:
Oh, and I'll update docs once this becomes a stable branch. Been over a month with no breakage, can you believe it???
Maintaining a fixed position isn't a minus, as translation tracking's an option, don't be sour-grapes-like.