Camera positioning

Cameras must be mounted firmly on tripods or other stable structures, which isolate the camera from movements or vibrations of any sort.

3D motion capture

To capture 3D data the camera system must consist of at least 2 cameras. The guidelines below can be used to set up a camera system.

The best possible setup for a 3D motion capture system is to position it so that all cameras can see the L-frame during the calibration, see chapter Calibration of the camera system.

The L-frame must be seen by at least two cameras. The other cameras must then have a sufficient amount of overlap between each others field of views for a successful calibration of the camera system. For this setup QTM will automatically use the Extended calibration method, see chapter Extended calibration.
Position the cameras so that there is as much possible overlap between the cameras view areas in all parts of the capture volume. Markers must be seen by at least 2 cameras to be able to calculate their 3D positions.
Try to maximize the angles between cameras for the best possible accuracy of the 3D data. The angle of incidence between any two cameras should ideally be more than 60 degrees, but at least 30 degrees. The accuracy of the 3D data calculated from only two cameras placed at less than 30 degrees can degrade below usable levels.
Try to avoid that cameras are in each others view areas. For example, place the cameras in a high position around the capture volume and point them down so that they are just outside of each others views.
Position the cameras so that their full sensor area overlaps with the capture volume. You can place markers at the edges of the capture volume to help positioning and pointing the cameras. Try to minimize the amount of unused sensor area.

In 2D motion capture just a single camera is needed, which is positioned perpendicular to plane that is captured.