Scroll Wheel Debounce Test

The scroll wheel debounce test records browser `WheelEvent` timing and the dominant delta axis inside the active card. It calculates adjacent event intervals, repeated same-direction pairs inside the selected rapid window, runs of closely spaced pairs, isolated opposite-direction events surrounded by matching neighbors, and a robust summary of delta magnitude consistency.

Those are software-visible patterns. Mechanical or electrical debounce happens inside the device or firmware before the operating system creates input messages. Standard browser APIs do not expose contact voltage, encoder transitions, firmware filters, or raw USB reports. For technical accuracy, the interface calls a flagged item a rapid repeated-event candidate or closely spaced event cluster—not a confirmed bounce event.

Conservative marks adjacent pairs at or below 18 milliseconds, Balanced uses 30 milliseconds, and Sensitive uses 50 milliseconds. Each preset also defines a wider neighborhood for checking whether one opposite sign sits between two matching signs. These values are user-selectable investigation heuristics, not manufacturer specifications or universal failure thresholds.

A sensitive preset will naturally report more candidates. A free-spin wheel, smooth touchpad gesture, momentum, rapid intentional movement, high event density, or browser coalescing can generate many close pairs without any defect. Start with a notched wheel and separated actions when possible. Keep the preset visible in any saved summary so two sessions are not compared as though they used the same rule.

A rapid same-direction pair means two adjacent dominant-axis events shared a sign and arrived inside the chosen window. Consecutive close pairs are grouped into a cluster so one burst is not mistaken for several independent episodes. An isolated opposite event is a sign reversal between matching neighbors within the preset neighborhood. It resembles a pattern users often investigate, but intent and momentum still matter.

Delta magnitude consistency uses the median absolute delta and median absolute deviation. A higher display percentage means magnitudes were closer to their median in this sample; it is not a health score. Different `deltaMode` units, acceleration, smooth scrolling, line settings, and input sources can change magnitude. Compare only runs made with the same environment and action style.

Place the pointer in the card, select Balanced, and make ten to twenty separate wheel steps in one direction. Pause briefly between physical actions rather than spinning continuously. Finish the run and read the timeline, where intervals inside the active close-pair threshold use a candidate style. Reset before testing the other direction so intentional reversals do not dominate the summary.

If the sample is too small, the tool refuses to offer a strong interpretation. If candidates appear, slow the action and repeat twice. Next compare the scroll jitter test, reverse test, and scroll jump test because they segment or calibrate direction differently. Agreement across controlled protocols is more informative than one candidate from a short mouse wheel debounce test.

One physical wheel action does not necessarily map to one browser event. Firmware, drivers, operating-system line settings, browser scheduling, acceleration, and smooth-scrolling layers can turn an action into a short event sequence. Trackpads and free-spin wheels commonly create dense tails. A wheel duplicate event test cannot recover the number of physical detents from the final event stream.

System load can also change interval spacing, and frame rendering is separate from input dispatch. Remote sessions, virtual machines, accessibility tools, and vendor profiles add more transformations. If a candidate cluster occurs only in one browser, investigate that environment before inferring a device issue. If it repeats across applications, document the input source and exact action rather than assigning a mechanical cause.

Confirm that the test direction was intentional and that touchpad momentum has ended. Try a simple document, disable only reversible application-specific smoothing or remapping, and compare a wired connection if the device supports one. Check battery and receiver placement for wireless input. Use official drivers and preserve current settings before changing them.

For external cleaning, disconnect the mouse and follow its manual. Do not blow debris deeper with unsafe tools, spray cleaner through the wheel opening, add general-purpose lubricant, or dismantle a covered device because the scroll encoder bounce test showed candidates. Persistent symptoms across systems belong in a manufacturer support or qualified repair conversation with the browser evidence clearly labeled.

No cluster proves an encoder fault, and no clean run certifies a mouse. Missing physical input may produce no event to analyze. The browser cannot distinguish a physical wheel, touchpad, trackball, remote input, or software-generated WheelEvent reliably. The test also does not measure USB polling rate or rendered scroll distance.

A responsible result says that repeated same-direction pairs, closely spaced clusters, or isolated opposite events were or were not detected under a named heuristic in the captured browser sample. Repeat the scroll wheel debounce test after one controlled change, compare another application, and avoid warranty or repair decisions based solely on this page.