C2.18.1multitouch contact limitdesignresearch

Hardware and the OS cap how many contacts can be tracked at once

Aliases: contact cap · ten-point touch · simultaneous contacts

What it is

Both the capacitive controller and the operating system allocate a finite number of contact slots. That simultaneous-contact limit is how many independent identifiers can be alive at once. Phones often claim five or ten; desktop monitors, automotive panels, and meeting boards differ. The cap is a tracking budget, not “how many fingers can touch the glass”: a finger that finds no slot is treated as absent.

Why it happens

Each tracked blob occupies filter state, coordinate history, and an id. The IC’s compute and SRAM are sized for that slot count; the OS input path has its own cap, and the smaller one wins. The grid’s ability to separate neighboring fingers tightens the effective count further: corners, wet skin, and a palm heel consume slots first, leaving fewer for fingertips. System gestures sometimes reserve slots for edge swipes or palm rejection, so the app’s usable number is below the headline. “Ten-point touch” is peak tracker capacity; live concurrency still subtracts occupied and unresolvable contacts.

Studying it

Add fingers one at a time past the claimed cap. Record whether each new contact gets a stable id, whether an old id is evicted, and whether report intervals stretch. Independent variables include landing spacing, presence of a palm heel, device orientation, and whether system gestures are enabled. Dependent measures are live identifier count, the moment an old contact loses its slot, and how recognizers fail near the cap. Treat the datasheet number as a hypothesis and count ids on the target device. Separate simultaneous ten-finger plants from sequential plants: the latter exposes first-come slot occupancy rather than instantaneous capacity.

Where it stops holding

Resistive, infrared-frame, and some optical touch systems derive their caps from different sensor geometry; a phone’s five or ten does not transfer. A stylus plus palm rejection may occupy a high-priority slot and leave fewer for fingers. The Web TouchList can be truncated again by the browser, so a native-app cap and a page cap need not match. Conductive gloves in the lab enlarge blobs and hit the separation limit early; that looks like a lower cap but is merging, not a full slot table.

Applying it

  • Discover the id count the app can actually track with a progressive-finger test on the target device, and write that number as a capability rather than copying “ten-point touch” from marketing.
  • Before shipping a three-or-more-finger shortcut, confirm foreground work, system gestures, and palm rejection will not exhaust the table.
  • Offer fewer-finger alternatives (a button, two fingers, a knob) near the cap; do not make a five-finger chord the default path.

Related

  • Same group: C2.18.2 Overflow contacts are dropped silently rather than signaled as errors · C2.18.3 On shared displays the contact cap bounds concurrent participants · C2.18.4 Contact limits vary by device generation and cannot be assumed uniform
  • Nearby: C2.19 Contact merging and splitting · C3.12 Three- and four-finger system gestures
  • Search terms: multitouch limit · simultaneous contacts · pointer id

Cards in the same group

Quick Actions

Share

Share this page

ios_share

https://hci.top/en/handbook/C2.18.1