zForce AIR Touch Sensor determine an object's position by signals derived from emitter-receiver pairs and have the capacity to detect and track several objects at the same time. Both the HW and the SW have been optimized in order to support standard touch gestures like, pinch-to-zoom, rotate, swipe and tap. However, some combinations of two or more objects might need special consideration.

Shadows

• An object directly behind another object cannot be illuminated.  In the figure above, object A will not be detected since illumination is blocked by object B.
• The correct receiver must have a clear field of view. Object B is in a region covered only by left looking receivers. Object B will not be detected because its field of view is blocked by object D.
• Object C may be seen by both left and right looking receivers. Although the right looking field of view is blocked by object D, object C is detected by the left looking receiver.
• Object D is detected by both left and right looking receivers.

Shadow Trick

Note that in most cases, user experience is not affected by the shadow situations mentioned above. This is because of a functionality implemented in the Touch Sensor firmware called "shadow trick", which e.g. generates a smooth "rotate" feeling despite one touch object being shadowed during the rotate gesture . A previously detected object that can no longer be detected is still reported as present if:

• The object was last seen close to a location where it could be shadowed by another object.
• The potentially shadowing object is still detected and hasn't moved away from a shadowing location.

The shadow trick make multi-touch gestures such as "rotate" and "pinch-to-zoom" work better. 

Adjacent Objects

• In order to recognize two objects close to each other (A and B), a separation must allow at least one emitter-receiver channel (~10 mm) to pass freely between them. Otherwise, the two objects will be reported as one large object.

More Than Two Objects

When more than two objects are being tracked the likelihood that an object ends up being in the shadow of another object increases. Therefore, it is only recommended to enable more than two tracked objects if, for example:

• it is not vital to track all detected objects 100% in all possible combinations and locations at all time. 
• When all objects are likely to be detected by the sensor, for example when it is expected that all objects will be placed along a line that is parallel to the sensor, as in the example below.