InputMetricInputMetric

Keyboard Chatter Test

Check for unusually fast repeated key presses that may match a double-typing symptom. The test separates held-key auto-repeat from new presses when the browser exposes enough information.

Start the test to focus this area

Normal held-key auto-repeat is ignored. Press Escape or Stop Test to leave test mode.

Keys Tested
0
Non-Repeat Presses
0
Suspected Chatter Events
0
Last Key
—
Last Press Interval
—ms
Shortest Suspected Interval
—ms
Most Affected Key
—

Recent Key Events

Key codeIntervalStatus
Check if each key registers
→ Keyboard Test Online
Compare key reset timing
→ Rapid Trigger Test

What the Keyboard Chatter Test Measures

Keyboard chatter describes an intended press that appears as more than one registered input. The tester compares new keydowns for the same physical key only after an intervening release.

The selected 20–50 ms threshold is a diagnostic filter. Normal held-key auto-repeat, marked by the browser's repeat flag, is excluded.

How to Use the Keyboard Chatter Test

  1. 1Start the test and leave the threshold at 40 ms initially.
  2. 2Press and release a suspected key deliberately, once at a time, for 30–50 attempts.
  3. 3Check whether the same key repeatedly produces suspected intervals.
  4. 4Try another key as a control, then repeat in another browser if unexpected.

How to Interpret the Result

A single fast interval is not a diagnosis; an intentional double tap can be flagged. A repeatable pattern on the same key during slow single presses is more useful evidence.

Compare connection mode or another computer if practical before investigating the switch or firmware.

Testing a Key That Types Twice

A useful run involves dozens of slow, deliberate taps on the suspected physical key, with a full release between taps. The recent-event table shows key codes and intervals, not a typed sentence. Try an unaffected key as a control so that a fast tapping habit is less likely to be mistaken for a device-specific pattern.

The browser's repeat flag distinguishes ordinary held-key auto-repeat from a new press in many cases. A second keydown without an observed release is not classified as chatter by this tool. Firmware debounce can also remove an unwanted transition before it becomes a browser event, so an empty suspected count does not rule out every physical issue.

Browser Limitations

The browser cannot receive raw switch voltage, sensor travel or events filtered by firmware. These intervals are between browser-observed key events, not contact-bounce duration. Physical key codes and timing stay local to the page.

How InputMetric measures browser input →

Related Tests

FAQ

Is chatter the same as normal key repeat?

No. Held-key auto-repeat is excluded when the browser marks an event as repeated.

What threshold should I choose?

Start with 40 ms. A higher threshold includes more intentional fast taps, while a lower one is stricter.

Does the test save what I type?

It keeps short-lived key-code counts and intervals locally; it does not build or transmit a typed sentence.