Inspect the interface the browser actually exposes
The event log is intended for developers, support conversations, and careful comparison. It preserves raw fields and labels units while avoiding unsupported claims about hardware distance or device identity.
How to use the Browser Wheel Event Tester
Start logging explicitly
Nothing is captured on page load. Press Start log and keep the pointer inside the capture area.
Exercise one behavior
Try a vertical wheel, horizontal gesture, or modifier combination separately so each pattern remains explainable.
Pause and export
Freeze the table, inspect relative timestamps and units, then copy JSON or download local CSV/JSON if needed.
wheel and scroll are different events
The browser dispatches a `wheel` event when it receives compatible wheel-like input. A `scroll` event occurs when the visible scroll position changes. A wheel event may be canceled, used for zoom, mapped to a canvas action, or received where no scrollable overflow exists. Conversely, a page can scroll through keyboard input, dragging, script, or accessibility controls without any wheel event. The wheel event tester records only the first part of that distinction.
This difference matters during debugging. If the WheelEvent inspector shows deltas but content does not move, inspect default prevention, overflow, event handlers, zoom gestures, and application mapping. If content moves without rows here, confirm that the pointer is inside the started capture card and that the movement was actually wheel-like input.
Read deltaX, deltaY, deltaZ, and deltaMode
`deltaX`, `deltaY`, and `deltaZ` are double-precision values describing components in the event’s coordinate axes. `deltaMode` gives their unit: 0 for pixels, 1 for lines, and 2 for pages. Raw values remain untouched in the wheel event log. A line or page value is never silently presented as a physical pixel distance, and a mode change is visible on the affected row.
Some interfaces estimate a display delta so unlike units fit one chart. Such conversion is necessarily heuristic because line height, viewport size, browser behavior, and application settings vary. This site keeps estimated display units out of the raw table and never uses them for a deterministic device comparison. Use the deltaY tester to compare patterns only within a controlled environment.
Modifier keys change how wheel input is used
The log includes Ctrl, Shift, Alt, and Meta state. Ctrl plus wheel is commonly associated with zoom, though exact handling is browser- and platform-dependent. Shift can map vertical wheel motion to horizontal movement in some applications. Alt and Meta may participate in shortcuts. A true value establishes only that the modifier state was present on the event, not what every software layer did with it.
When debugging, test one combination at a time and avoid browser shortcuts that could navigate away. Compare the delta axes and mode before and after holding the modifier. The wheel event tester intentionally does not capture keyboard events separately or infer a shortcut beyond the modifier fields already present on `WheelEvent`.
Input source and raw-value limitations
A mouse wheel, touchpad, trackball, accessibility device, remote session, or software-generated input can all reach the same interface. Standard WheelEvent fields do not provide a dependable device model, encoder type, physical notch count, DPI, or hardware polling rate. Patterns can suggest a smooth or step-like source, but the WheelEvent inspector labels no event as a certain device type.
Raw deltas can change with drivers, natural scrolling, acceleration, line preferences, zoom, browser version, and application logic. Cross-device benchmarks based on those values can therefore be misleading. Keep the full environment fixed when reproducing a front-end bug and include unit mode and relative timing with any shared sample.
Listener behavior, performance, and privacy
High-frequency event handlers should do little synchronous work. This tester buffers events and updates the interface in animation-frame batches, retains only 500 rows, and drops the oldest row after the cap. The table can be paused without clearing data. No document-level listener is installed, so page scrolling outside the capture area remains available.
A passive listener cannot call `preventDefault`, while a non-passive listener may cancel default behavior when a focused interaction genuinely needs it. This general event log does not cancel vertical scrolling. Horizontal containment is limited to its dedicated test. All table rows and exports are built locally; the log excludes screen coordinates and does not include identity fields, accounts, or analytics identifiers.
Browser Wheel Event Tester questions
Why is deltaY a decimal?
WheelEvent deltas are floating-point values. Smooth input, scaling, and software processing can produce fractional values.
What happens after 500 rows?
The oldest row is discarded as a new one arrives. The visible notice explains when the bounded log is at capacity.
Can the WheelEvent inspector tell mouse from touchpad?
Not reliably. Standard fields describe the event, not a trustworthy source-device category.
Are relative timestamps personal data?
They are elapsed times within the current session and are processed locally. The export contains no screen coordinates or identity field.