Scroll Wheel Sensitivity Test

Scroll sensitivity can mean an operating-system line setting, a vendor utility option, application scaling, or simply how strongly and quickly a person moves the wheel. A web page cannot read most of those settings independently. This scroll wheel sensitivity test therefore measures mouse wheel response at the browser boundary: the event count, density, timing, delta values, direction signs, and units produced during three controlled input stages.

The wheel sensitivity test does not measure mouse DPI. DPI describes pointer motion resolution and is unrelated to a standard WheelEvent delta. It also does not discover a device model, firmware value, physical notch count, or vendor sensitivity number. The useful question is narrower: under the current environment, how does the browser-observed response change when you intentionally perform slow, normal, and fast scrolling with one protocol?

Guided actions asks for a chosen number of separated movements in each stage. An inactivity gap segments the stream so the result can report median events per action. This mode is useful when you can make deliberate strokes and want group-level context. Timed run instead captures everything during a fixed interval. It reduces the need to count actions but reflects how much input you produce in that time, so technique remains part of the result.

Slow should be comfortable and deliberate, Normal should resemble everyday use, and Fast should be controlled rather than violent. The labels are instructions, not calibrated speeds. Complete stages in order and avoid practicing one stage far more than another. If you lose focus, use a modifier, change a setting, or reverse unintentionally, reset the protocol. A repeat made with cleaner technique is more informative than explaining away a compromised stage.

Event count is the number of messages received. Events per second divides that count by active run duration; it is browser event density, not hardware polling rate. Median event interval describes typical spacing. Median and p95 absolute delta summarize the center and upper edge of raw magnitudes, while cumulative absolute delta adds magnitude within that stage. Guided mode also shows median events per segmented action.

Direction consistency is the share of nonzero samples using the dominant sign. It helps reveal whether a supposedly one-direction run included reversals, but natural-scrolling preferences determine how sign relates to physical motion. The deltaMode row identifies pixel, line, or page units. When stages use different modes, the tool disables raw magnitude comparison while retaining event count and timing. Converting incompatible raw units into one score would imply precision the browser does not provide.

After completing three stages, Save as Baseline A stores their summaries in this browser's local storage. You can then change one operating-system or official mouse-software setting, repeat the same protocol, and treat the new three stages as Run B. The page checks deltaMode compatibility before describing raw magnitude as comparable. Timing and counts remain visible when units differ, with a clear limitation message.

The settings comparison never reads the setting you changed. Label your process outside the page if necessary, alter only one variable, and restore it manually when finished. Local Baseline A is not a cloud profile, universal calibration, or fingerprint. Another person, computer, browser, device, direction, or technique creates a different context. Use several A/B pairs before deciding that a setting caused a stable response change.

A user adjusting line count can document whether browser event magnitude or density changes. A gamer can compare a binding response before and after an official profile adjustment without confusing it with pointer DPI. Developers can examine how a canvas or list receives slow and fast wheel input. Support staff can request a fixed protocol when scrolling feels too sensitive. Accessibility users can compare remapped or assisted input while keeping the limitations explicit.

The scroll response test is also useful alongside the scroll speed test, which focuses on fixed-time rate; the free-spin test, which focuses on momentum tail; the step test, which groups isolated actions; and the comparison test, which locks a broader A/B protocol. Choose the tool that matches the question. More metrics do not make an uncontrolled comparison valid, and no universal sensitivity score applies across browsers.

Device mechanics, firmware, connection, driver transformations, operating-system line preferences, natural scrolling, browser version, zoom, application behavior, system load, and hand motion all affect the mouse wheel response. Background tabs may be throttled. Touchpads and accessibility devices can reach the same interface. The scroll wheel sensitivity test cannot isolate these layers or determine why a stage changed, so its results support controlled troubleshooting rather than diagnosis.

Keep the pointer inside the active area and the tab visible. Use one direction and similar hand position. If guided actions merge, pause longer. If a stage has no events, verify focus and use the wheel event tester. If deltaMode changes, finish for timing information but repeat before comparing magnitude. When results differ across applications, review each application's scroll behavior and official driver profile. Do not disassemble or clean internally based on this test.

The three active event streams are processed in memory and summarized locally. The optional save control stores a versioned result on this device, while Baseline A stores only stage summaries. The test does not upload a sensitivity profile, inspect installed mouse software, automatically detect OS settings, or infer a device model. Reset clears the active comparison so you can start a clean protocol.

Results describe browser-observed scroll response under the current device, settings, operating system, browser, protocol, and user technique. They are not a DPI measurement, physical calibration, manufacturer specification, or definitive hardware diagnosis. Read the testing disclaimer before relying on a difference for support, repair, warranty, competition, or purchasing decisions.