The interaction volume is the subspace of capture range treated as valid input

What it is

The interaction volume is the subspace of capture range that the product treats as “motion here counts as input.” The sensor may still track a hand outside that region; the application layer treats those points as invalid. It is a software policy: a box in front of a screen, a shell in front of the body, or an XR near-field that travels with a window. Capture range answers “can it be seen”; the interaction volume answers “if seen, does it count.”

Why it happens

Implementations typically take tracked coordinates and run a containment test: only points inside the box enter the gesture state machine. The box may be a fixed depth extruded from the display plane, or an offset from the user’s torso. Because it is policy, the volume can change three times a day without touching hardware: an exhibit shrinks it so aisle traffic does not fire, a living-room TV grows it so the sofa can reach. It can also switch per application—browsing uses a larger volume, payment confirmation shrinks to a narrow cone facing the screen, cutting bystander entry at the source. A product with no explicit interaction volume treats the entire capture range as valid input and turns wherever the hardware happens to see into an operable surface.

Studying it

On a site whose capture envelope is already measured, mark the six faces of the software box. Have participants perform the same gesture inside the box, outside the box but inside capture range, and outside capture range. Record whether the system responds, whether a tracking visualization still shows the hand, and how users explain non-response. Vary box size and placement and measure time-to-arrive, false triggers, and “I thought it would work here” reports. The point is the behavioral difference among the three places, not a single recognition rate.

Where it stops holding

In a headset the “interaction volume” is often just the near field the wearer’s hands already reach, a different geometry from the fixed box in front of a desktop camera. Full-body games that treat the whole room as the interaction volume lose the motive to shrink a subspace; the problem becomes room-scale calibration. If the volume follows a body anchor, it is no longer furniture in the room, and measurement has to move into torso-relative coordinates. Drawing an interaction volume larger than capture range is a no-op: the extra region will never have signal.

Applying it

• Configure the interaction volume as geometry independent of sensor model. When the installer swaps hardware, remeasure capture range; do not store both numbers in one field.
• In the debug view, draw the capture envelope and the interaction volume in different colors so engineers do not equate “can track” with “can operate.”
• Write copy as “operate in this region in front of the screen,” not as the sensor’s maximum distance. That number is capture range; users cannot use it.

Related

• Same group: C4.20.1 Capture range is the volume in which the sensor can physically detect motion · C4.20.3 Motion inside capture range but outside the interaction volume is easily mistaken for system failure · C4.20.4 The interaction volume is typically smaller than capture range to leave a margin for tracking jitter
• Adjacent: C4.21 Coordinate anchors of the interaction volume · C4.10 Sensing volume and interaction distance
• Search: interaction volume · valid input volume · software bounds