Testing Methodology
Understand the browser signals behind each I/O Checkup hardware family, how to compare results, and the limits of online diagnostics.
What an online check can tell you
We use browser events, browser media APIs, generated sound, visual patterns, and printable references. These provide observations you can repeat. They cannot inspect hidden firmware, certify calibration, or establish the cause of every hardware fault.
Keep the browser, connection, device position, and surroundings consistent for a baseline. Repeat a suspicious result, then change one variable at a time. Compare with the manufacturer diagnostic before making a repair or warranty decision.
Keyboard checks
Press keys systematically and compare the browser event log with the physical layout. Timing reflects event delivery and browser workload, not a laboratory measurement of switch-to-screen latency.
Mouse checks
Clicks and scrolling are counted from browser events. DPI is estimated from pointer movement over a distance you measure. Acceleration, scaling, connection, and technique can affect the estimate.
Display checks
Patterns provide visual references for your own inspection. Refresh estimates come from animation callbacks, not direct panel telemetry. Lighting, display processing, viewing angle, and browser behavior affect results.
Touchscreen checks
Contact positions, identifiers, and pressure come from Pointer Events where available. Pressure may be a default value on unsupported hardware. Browser gestures and palm rejection can interrupt contact tracking.
Gamepad checks
The Gamepad API supplies button values, axes, and timestamps. A trigger value is normalized input, not physical pressure. Timestamp changes are browser-visible updates, not proof of a hardware polling specification.
Microphone checks
Web Audio analyzes the microphone signal after you grant access. dBFS describes digital signal level; it is not calibrated sound pressure in dB SPL. Gain, noise suppression, distance, and room noise affect readings.
Webcam checks
Video track settings describe the negotiated stream. Reported frame rate is not a guarantee that every frame was delivered. Brightness and color readings summarize the preview rather than certify exposure or color accuracy.
Speaker checks
Web Audio generates tones for listening checks. Start at low volume and stop if uncomfortable. Audibility depends on the output device and hearing; these tools are not medical hearing tests or calibrated frequency-response measurements.
Printer checks
The browser creates reference artwork and opens the print dialog only when requested. Inspect the physical sheet at actual size. The site cannot read ink levels, queues, mechanical condition, or whether the job completed.
Compatibility and review
A supported API does not mean every device exposes every measurement. Unsupported features should be shown as unavailable, not as a failed device or a zero reading.
Our release checks use automated route and layout tests alongside targeted browser interactions. A simulated microphone, camera, or touch event checks the software path, not your physical device. We do not claim laboratory validation or universal browser compatibility.