[ANN] opentrack 2.0 beta 1 released!
-
Try using .bmp as the extensions instead of .jpg. JPEG support hasn’t been compiled in that particular opencv build. It should work on Windows XP/2003 and above with the file format caveat.
PT’s coplanar too, since it uses 3 points. That amount forms a plane, always. The real issue is perpendicularity toward the camera.
In a yet-unpushed build, settled for iteratively improving the previous frame’s pose, until the pitch sign changes. This has its own caveats, but completed a test flight without a hitch.
Changing the head centroid works too, and yet-unpushed change allows for skewing the 3D coordinates of the marker’s edges on the pitch axis, to account for the marker’s skew. This improves on head centroid position.
Finally, distortion coefficients are necessary to prevent rot/t interconnect. This is an issue given that barely any users care to calibrate the webcam model they’re using.
-
In case there are any outstanding issues with regard to the software, please report them. They’ll be corrected by me when it’s possible.
It’s been a while, after all. Happy head-wiggling!
-sh
-
Still using it and happy with everything.
Question though, any plans to add DX centering/pause commands to it?
-
In case there are any outstanding issues with regard to the software, please report them. They’ll be corrected by me when it’s possible.
It’s been a while, after all. Happy head-wiggling!
-sh
Please take a look at this issue:
https://github.com/opentrack/opentrack/issues/25 -
btw. the windows binaries are down:
http://sourceforge.net/p/facetracknoir/discussion/1150909/thread/fc841d30/ -
Here’s what seems to be the latest binary:
https://www.dropbox.com/s/j8ygiiq5ex6dapi/opentrack-win32-portable-2.2-20140116.tar.xz
As for DX centering, the idea’s to offer the same core functionality on all supported platforms. Also, it’d be relatively much work.
If you need to center THAT much, it’s probably an issue in itself, and should be addressed instead of working around it.
-
Latest versions act weird on my PC. The view is skipping back and forth when i pan, as if it was sending positions in the wrong order. Can’t explain it, here’s a video:
Doesn’t seem to be Falcon-specific, it does same thing in X-Plane and Euro Truck Simulator.Tested it with binaries from 2.2-2014011 to 2.2-20140116 and it happens with 2.2-20140113 and newer, previous versions work fine.
EDIT: it also seems to get worse with higher framerate.
-
Did work on it yesterday and today. Try today’s build instead. Several users reported the issue and it’s apparently better now. If not, can analyze the code some more.
In tomorrow’s build, EWMA filter becomes actually useful.
Great job bisecting the issue.
-
This will probably sound dumb, but i have no clue where to download today’s build. Newest binary in dropbox is 6 days old (bug still present), only file from today is a reupload of 20140116. Compiling it from source is way beyond my skill level.
-
Deleted the wrong things during cleanup. Try now.
-
No glitches so far, appears to be fixed. Thanks. Going to test it with other games too.
EDIT: X-Plane OK, ETS2 OK
-
Hey sthalik,
Just FYI, if you dont have Opentrack running, you cannot use your mouse to pan around the pit in BMS. The only workaround is to delete the Opentrack folder, or to run BMS from a new windows session. I know a few users who had problems and even tried reinstalling BMS, all for nothing.
The issue is not with BMS itself : with a Track IR, even with the soft running but not actually tracking, the mouse works fine.
So with OpenTrack, BMS still detects a head tracking device even though it is not running. Any chance you know what is going on ?
-
It’s the “NPClient.dll” file that needs renaming/deleting, rest of it is fine.
You can also disable "enable 3D TrackIR and it’ll work like that, too.
As for a proper solution, the NPClient component now doesn’t check whether there’s opentrack running and assumes it is. Maybe it can be fixed in time for 2.3, open a bug tracker issue so it doesn’t get lost.
-
i have started using my xperia z1 as a tracking device. it works great
not quite free but if you don’t want to build something or have bad results
with face tracking i highly recommend using a smartphone
-
i have started using my xperia z1 as a tracking device. it works great
not quite free but if you don’t want to build something or have bad results
with face tracking i highly recommend using a smartphone
What are your settings do in opentrack? I have an Huawei P7 , it could be working with it?
only 3dof or 6dof is enable too -
What are your settings do in opentrack? I have an Huawei P7 , it could be working with it?
only 3dof or 6dof is enable toopitch, roll and yaw work only. i downloaded freepie miu onto my phone and played about with my curves works great
-
THX! It’s works . I found an another way about smartphon…
you can use the smartphone as occulus rift with almost every PC game
http://www.reddit.com/r/Android/comments/2bo4jx/google_cardboardbased_diy_rift_works_with_amd/I didn’t try this yet, but if someone try it , It would be good to write some experience about it. Which config, which set up.
-
New stuff inbound.
The biggest change pending release is dynamic point detection. Specify point size and they’ll get detected irrespective of lighting. Known as “automatic threshold” in PT options. The effort was sponsored by TrackHat. Furthermore, on false detections, more circular points are used, ignoring rectangular false detections and so on.
Another change is return to Accela roots of mk2 version. There are some issues to be sorted out and that’s holding up the next binary release.
PointTracker now doesn’t yaw on pure X translation, and so on.
We’re getting pretty stable and fun after X release candidates, there are no more last-moment changes anymore that violated the release cycle.
-
Thank you sthalik, and good to see you back.
Vags
-
is there an exe or an installer for this?