Is your screen, camera and mic actually working?

In full screen, click or press any key to move to the next colour and Esc to exit. Black and white reveal most faults; red, green and blue isolate a single stuck sub-pixel. The gradient exposes banding and uneven backlight.

Check: does the image freeze or smear when you move? Is text on a page held up to the lens readable? Many laptop cameras are 720p — the resolution readout above tells you the truth.

Speak at normal volume from where you actually sit. The level should bounce in the green-to-amber range; pinned at max the whole time means distortion, and barely moving means the wrong mic is selected — headsets and webcams often add their own.

Don't forget the easy-to-miss ones: function row, Caps Lock, both Shift keys, Enter, and the space bar pressed at its far left and right edges — a space bar that only works in the middle has a failing stabiliser or switch.

With headphones, left-only should be in your left ear only. If both tones land in both ears, your output is set to mono somewhere (system accessibility settings are the usual culprit). If the 80 Hz tone is silent on laptop speakers, that's normal — small drivers can't reproduce it; on full-size speakers or a subwoofer it should be clearly felt.

Watch the two blocks cross the strip: the half-rate block steps and judders, the full-rate block glides. If both look identical and the reading says 60, your display chain is set to 60 Hz — the test is showing you the truth, not the box spec.

To check for unwanted double-clicking (“chatter”), click the left button once, slowly, ten times, and compare with the counter: if the count runs ahead of your clicks, the switch is bouncing. For movement: sweep the pointer steadily across the pad — the distance and speed readouts use the raw movement your mouse reports, which is also how you'd spot a sensor that skips or stalls on fast flicks.

For a fair result: pause downloads and streaming, and test twice — once on Wi-Fi where you normally sit, once with a cable into the router (or standing next to it). If the wired result matches your plan and the Wi-Fi result doesn't, your provider is fine and the wireless link is the problem. Figures are the real transfer measured by this browser, not an estimate.

Handy pasted into a support ticket: it answers OS, browser, screen and GPU in one go. Nothing here leaves the page unless you copy it.

Drag one finger slowly along each edge and through every corner — edge and corner dead zones are the most common touchscreen fault and the easiest to miss. Then check multi-touch: most modern screens track 5–10 points, so if yours stops at 2, something is limiting it.

Work through every button once, roll both sticks in slow circles, and squeeze the triggers gradually — the axis bars should move smoothly and, importantly, come back to exactly 0.00 when you let go. A stick that sits at 0.10–0.20 untouched is drifting.

Smooth gradients should shade evenly with no visible steps or colour fringes. Banding shows as distinct vertical bands instead of a continuous ramp — some banding is normal on 6-bit laptop panels, harsh steps are not. Black level: you should just be able to separate the lowest squares (2–5%) from pure black. White saturation: you should just be able to separate the highest squares (250–254) from pure white.

The beat test works because your tap has to wait for the sound to reach you: if sound leaves late, your taps land late by the same amount. Ignore the first beat, tap the remaining ones, and use wired headphones or speakers for the baseline — then repeat on Bluetooth to see exactly what it costs you. Typical results: wired ≈ under 30 ms of tap delay beyond your reaction pattern, Bluetooth ≈ 100–300 ms.