Hi Michael:

I emailed you the scene.xml file that I used as my first estimate. You can see that some of the cameras align ok against the calibration video and it can give you the layout of the cameras in the room. If you need more precise measurements, I'll have to do some calculations to make a better estimate.

Much obliged,

Michael 2

Hi Michael:

OK, I was able to finish the tracking of the dance sequence with the new calibration you sent me. DMC seems to have some problems tracking my dancer whenever she's got her shoulder twisted in an extreme position. Dancers are very flexible, so their joint constraints should probably be less confining than for normal people. I don't know if DMC has the potential to change the joint constraints, but it would be helpful.

It would take a ton of one-frame-at-a-time manual tracking to fix the bad arm tracking, and I find the manual tracking inefficient in DMC. The main problem I noticed is that I'm having to click in one part of the screen to advance one frame, and somewhere else to pick the editing tool (IK, rotate, etc.) and then click back inside the video to make the change followed by clicking at the top of the screen to chave camera views. It would be helpful to have some way to advance/reverse the frame, pick the tool, and cvhange cameras without moving the mouse pointer away from the figure being worked on. Professional 3D/CAD software packages use multibutton mice to accomplish these tasks, and I'm guessing that something like this would be the way to upgrade DMC.

I will be using the maglight for calibration next time. In addition, I'm going to experiment with camera placement, but all the configurations will have all the cameras pointing towards the center of the room.

Thank you again for your exceptionally helpful customer support,

Michael 2

Hi folks:

I just had another mocap shoot. I corrected all the isses that Michael Nikinov identified, and was able to get a very good calibration. I tried out 3 different camera placement configurations, and the one I just calibrated has 4 cameras at 10 foot high in the corners of a room which is 21 x 24 feet. The other two cameras are on normal tripods about 4 feet high, and they are in opposite diagonal corners of the room just below the 10 foot high cameras. The tracking seems to be a lot better than the previous attempt, so I'm feeling very hopeful at this time.

We'll see how well things have baked when we open the oven in the morning. Thanks to everyone for their help, but especially to Michael for helping with the calibration.

Regards,

Michael 2