Input Device Diagnostics Studio

It is a web-based testing utility that checks whether your browser accurately registers physical input events (like keydown, pointerdown, and gamepad axis changes), visualizing them on screen in real time.

Ensure the test area on the webpage has active focus by clicking inside it. If specific keys still fail to register, they may be intercepted by OS-level shortcuts (e.g. Cmd+Space or Alt+Tab), or there may be a physical switch contact issue.

'code' represents the physical key position on the keyboard (e.g. always 'KeyQ' regardless of layout), whereas 'key' represents the resolved character or function produced, which depends on active language layouts and modifier states.

Location distinguishes keys that share names: 0 is standard, 1 is left modifier (e.g. Left Shift), 2 is right modifier (e.g. Right Shift), and 3 represents keys on the numeric keypad (Numpad).

It tests how many keys pressed concurrently can be received by the browser at the exact same moment without dropping signals.

Rollover refers to how many concurrent key presses a keyboard can register. This tool displays 'browser-observed simultaneous keys', which reflects browser and OS event receipt rather than proving an electrical hardware N-key rollover specification.

Ghosting occurs when pressing three or more keys on an unoptimized matrix circuit causes unpressed keys to trigger falsely or intended keys to drop. You can test key combinations here to verify if the browser receives the expected keys.

Chatter happens when a degraded or oxidized physical switch contact causes a single keystroke to trigger multiple rapid down/up signals in a few milliseconds. We flag repeated events occurring within <= 35ms.

When holding down a key, the OS and browser fire continuous repeat keydown events. This tool measures the initial delay before repetition starts and the mean interval between repeats.

Yes. Left click (Button 0), Middle click (Button 1), Right click (Button 2), and accessible side buttons (Button 3: Back, Button 4: Forward) are tracked individually.

Yes. The tool measures elapsed time between consecutive clicks in milliseconds to help detect faulty switches that trigger accidental double-clicks.

CPS measures clicking speed. You can start a timed 1s, 5s, or 10s test to measure browser-observed clicks per second.

Yes. Vertical (deltaY) and horizontal tilt (deltaX) wheel motion, along with deltaMode (pixels, lines, or pages), are tracked in real time.

Yes. Accessing this tool from a smartphone, tablet, or touchscreen laptop lets you observe touch contact points, coordinates, and contact pressure.

The tool visualizes all concurrent touch points reported by the device hardware and browser (e.g. 2, 5, or 10 fingers) and logs the peak count.

Yes. The engine automatically analyzes finger trajectories to identify swipes, long presses (>= 500ms), two-finger pinch-to-zoom, and rotation angles.

Yes. Using the standard W3C Gamepad API, USB and Bluetooth controllers are automatically detected without requiring software installations.

D-pad, A/B/X/Y, bumpers (LB/RB), analog triggers (LT/RT from 0.00 to 1.00), stick clicks (L3/R3), Start, Back/Select, and dual analog stick axes (-1.00 to +1.00).

Stick drift occurs when an analog thumbstick registers movement even when left completely untouched in its resting neutral position.

A dead zone is a threshold region around the center where small stick movements are ignored. The tool measures neutral resting deviation to evaluate effective dead zones.

No. Browsers receive events through multiple abstraction layers—operating systems, device drivers, and browser event queues. Browser observation cannot prove electrical hardware failure, but it provides objective empirical evidence of observable behavior.

No. Gamepad ID strings are displayed on screen for user convenience only. They are never sent to analytics, stored in cookies, or used for fingerprinting.

Never. All event handling is 100% client-side in transient memory. There are zero network requests, zero server analytics, and zero cloud uploads.

No. We never use localStorage, sessionStorage, IndexedDB, or cookies. Closing or refreshing the page immediately purges all recorded events from memory.

Different browsers have subtle variations in Web API event timing, task scheduling, compositor refresh rates, and OS shortcut handling.

For user privacy, Firefox requires pressing a physical button on the gamepad before exposing connected controller information to the page.

No. Browser JavaScript is constrained by display refresh rates and task scheduling. Measured frequencies reflect browser event loop sampling rather than raw hardware polling rates.

No. The browser can only measure relative time from when an event enters the web runtime, not the hardware transit time from physical switch actuation to the OS driver.