Huge calibration / calculation problems with 12 camera setup

Post Reply
onkelandy
Posts: 3
Joined: Thu Dec 10, 2015 5:46 am

Huge calibration / calculation problems with 12 camera setup

Post by onkelandy »

Hi all!

I've searched on the forum and hoped to find a solution for our problems but didn't. We've been trying to calibrate a 12 camera setup with Flex 3 cameras and ARENA 1.7.3. Nothing seems to work.
- Rectangular Setup 6x10m, as described in the tutorial video; upper cameras portrait, lower landscape; all landscpae; all portrait
- "Circular" Setup: no perfect circle, but diameter of about 5-6m; 6 tripods with 2 cameras, all combinations of portrait and landscape mode
Camera height: lower between 1,7m and 2m (changed that during several setup tests), upper cameras between 2,4m and 2,8m.

PROBLEM: Even though the cameras are capturing a lot of samples and tell us the data is good enough for "High Quality" calculation the calculation process itself is not successful. Even after one hour the quality is poor. As far as I understand it should change to medium after about 10 minutes.. our PC is an HP workstation, so fast enough I guess.

We're using the small 3 marker wand and have the correct standard settings in Arena.
VERY weird: As we made the rectangle smaller, about 6x9m we had instant success, perfect calibration and calculation. After wanding once more the next day we already had the same problems as always - meanerrors of about 10 on all cameras, no useful calculation. Although we didn't change anything in the setup :(

We've also tried to aim the cameras differently. Starting from point all to the middle of the stage and adjusting the cameras so the 1,5m high marker is in at the center of all cameras. No success. Aiming upper cameras so that they just don't see the opposite cameras and aiming the lower that the 1,5m marker is just "at the top" of the image, etc., etc. We've tried almost everything now. Besides the one weird insidence described above we've also setup the system successfully several months ago and also years ago. So the cameras should be okay, they also sync as they should..

And the wanding results usually quite good. About 1000 samples per camera, covering the whole camera space with green lines.. But calculation still takes forever :(

Hope anyone can help.
steven.andrews
NaturalPoint Employee
NaturalPoint Employee
Posts: 738
Joined: Mon Jan 19, 2015 11:52 am

Re: Huge calibration / calculation problems with 12 camera s

Post by steven.andrews »

Hello onkelandy,

I'm sorry to hear about the problems you have been experiencing with your camera calibrations. I would live to look into this with you to see if we can determine what is causing your poor calibrations.

Could you please open a support ticket with us at help.naturalpoint.com? I would like you to share a recording of your wand wave. This may require you to initiate a recording and performing a mock wand wave.

Cameras should be in one of the standard rectangular or circular configurations for this test.

Best regards,
Steven
--
Steven Andrews
OptiTrack | Customer Support Engineer
onkelandy
Posts: 3
Joined: Thu Dec 10, 2015 5:46 am

Re: Huge calibration / calculation problems with 12 camera s

Post by onkelandy »

Thanks for the reply. I've opened a ticket and attached some wanding data.
onkelandy
Posts: 3
Joined: Thu Dec 10, 2015 5:46 am

Re: Huge calibration / calculation problems with 12 camera s

Post by onkelandy »

We are very glad that thanks to a remote session we could fix the problem.
After changing the focus of most of the cameras so the markers don't have any halo effects anymore everything works fine!! Thanks alot.
Post Reply