Real signals. Clear limits.
Our test pages use standard browser capabilities. We distinguish measured events, reported capabilities and observations that only you can confirm.
Measured in the browser
Keyboard and mouse tools count browser events. Touch tools receive pointer events. Audio input tools inspect local waveform levels. Webcam tools display captured video and report browser-provided dimensions. These are useful observations, not factory calibration.
Reported by browser APIs
Capabilities, permissions, battery and hardware identifiers depend on the browser and OS exposing an API. "Unsupported" does not mean a device is physically broken. Privacy settings may deliberately hide information.
Visual and audible checks
Speaker, stereo, contrast and dead-pixel tests show or play patterns. Only the person using the device can say whether a speaker sounded correct or a pixel was defective. Start speaker checks at low volume; avoid prolonged loud tones.
Timing estimates
Frame rates are estimated from animation scheduling; network timing measures response to this website, not end-to-end broadband speed. Background tabs, power saving, browser throttling, network routing and caching can affect measurements.
Practical repair instructions
We prioritize reversible checks: permissions, cables, settings, restarts, alternative apps and retesting. Follow vendor-specific documentation for physical disassembly or firmware work. Stop if a device is swollen, wet, hot or electrically unsafe.
How we maintain accuracy
We test core interfaces in automated browsers and include clear limitations on each tool. Real physical hardware and operating-system combinations still require on-device testing. Reported outcomes should be interpreted in that context.
Ready to test? Open the diagnostic bench ↗