Vector Expansion

[Bastich]

Hello

We purchased 3 of the "Opti Track" systems to evaluate and see if they can be integrated into our current VR system. It would be nice to get rid of the magnetic tracking system as this is a much tidier and far more cost effective.

The applications we produce are written in C++ using DirectX 8.1. The VC6 example was fairly useful however I have removed MFC as it is not needed or wanted.

I have had basic dot tracking up and running stably for a while now. However soon after initialising the "Vector Expansion" (5 - 10 minutes) the system promptly crashes. The "Call Stack" from VC6 at the point of crash is:

OPTITRACK! 019eb053()
OPTITRACK! 019eb3a9()
OT_Camera::Update_Vectors() line 232 + 56 bytes
OT_Camera::Update() line 153
OT_Control::Update() line 419
App::Update() line 170
CDevice::Update() line 580 + 13 bytes
CGFX::Device_Update() line 95
CSys::Run() line 515
WinMain(HINSTANCE__ * 0x00400000, HINSTANCE__ * 0x00000000, char * 0x00141f16, int 1) line 198
WinMainCRTStartup() line 198 + 54 bytes
KERNEL32! 7c816d4f()

Has anyone got any suggestions as it appears as if it runs off the end of a list or similar (delayed crash). Other than that what resources apart from the "camera" frame do I need to release every main loop?

Please help

James Roberts

Head of Software Design and Development

http://www.Maelstrom.com

[Birch]

James,

Would it be possible to send us a copy of the modified source you are working on? It would make it easier for us to do the debugging work.

Also, if you can send us a screenshot of what the camera data looks like when you are testing, that will help reproduce the operating conditions.

If you can send us these, then please email them to support@naturalpoint.com with a note that it should be routed to Birch and Eric.

Thank you.

[Bastich]

Thanks for the response.

How much of the code do you want as I can send you a copy of the current VC6 Test Rig project (2meg RAR file). If so you will need to have DX8.1 SDK and TokaMak physics libraries installed.

Alternatively I can send you just the cpp / h files for the tracker integration (adapted from example project file)

Let me know which option would be best for you and where I can send the relevant files to.

Thanks

James

[ericd]

It's probably better if you could send the whole rar file over to make sure that we have as close to the same environment as you do to reproduce the problem. If you could email the file to ericd@naturalpoint.com I'll take a look at it.

I grabbed the physics library so that shouldn't be a problem.

Thanks!

[Cristian]

Hi Eric and James,

Have you succeeded in removing the MFC from the sample code?

Please could you post it somewhere so we could download it?

Thanks a lot!

[buw]

I'm interested too...
That would be really nice...

[Bastich]

Finally got back to the integration of the Opti Track system (Been busy on a 4 player networked VR Sky Diving game). Unfortunately I still seem to get the same crash, ether in camera->update or vector->update. The crash occurs anywhere from 1 minute up to many hours (left over night on a number of occasions). I have stripped all the MFC from the example tracker code and wrapped it into a pure C++ Win32 app. Once the current crash has been fixed I will be pleased to send the code to any other interested parties.

Would it be possible to send my app in again so that you could run it in debug and possibly spot what�s going on?

[jspoxon]

Hello I am new to this forum,

I am interested in trying to get Track IR working without the MFC. I will probably dedicate some time the next couple of days. However, should you fix the problem I'll be glad to get a copy of that app you mention.

Cheers,

Jaime

[bierz]

Iwould also be interested in an example source code without the mfc

[Bastich]

I will give the none MFC code to anyone interested once the crash bug is fixed. However I posted a request for help last September and have had absolutley no response.

James Roberts

Head of Software Design and Development

http://www.Maelstrom.com