Free · Runs in your browser · No install

One toolbox,
for everyday checks and conversions

Twenty-odd tools in one place: test mouse double-clicks, check left/right audio, find dead pixels, convert an image, run a few numbers. Open it and use it — nothing to install.

Browse all tools Test my mouse →

Stays on your device by default
No sign-up
Chinese & English

SIGNAL SCOPELIVE

↑ Illustrative waveform, not a measurement — the real test is the button below

LATENCY— ms

JITTER— ms

STATUSIdle

Pick a category and start

Everything runs in your browser — without an account, no test data is uploaded. Numbers are reported as far as they can be trusted, and their limits are stated.

10 toolsInput devicesMouse buttons, wheel, polling rate, DPI, acceleration, and per-key keyboard response.MOUSE · KEYBOARD · POLLING 2 toolsDisplayDead pixels, bright spots, backlight bleed, and a visual refresh-rate check.DEAD PIXEL · REFRESH 2 toolsAudioMicrophone input, speaker channels and sweep — with a note on what the browser changes.MIC · SPEAKER · SWEEP 1 toolCameraActual resolution, frame rate and colour cast — read live from the stream.CAMERA · RESOLUTION · FPS 9 toolsUtilitiesScreen recording, image conversion, favicon generation, hashing, JSON, QR codes, unit conversion.CONVERT · HASH · ENCODE 5 toolsPracticeReaction time, aim, steadiness, doodling and sensitivity conversion.REACTION · AIM · STEADY

All tools

Grouped by what they do. The badge on each card says how far the numbers can be trusted.

Input devices

10 tools

Buttons / wheel / hold / double-clickFour readings from one pointer stream: which button, how long, double-click gap, wheel direction.Measured directly CPS click speedClicks per second, with the average and the best 1-second window.Measured directly Mouse trailRead the shape: isolated spikes mean dropped frames; a sustained change means you moved faster.Shape only Polling rateReports per second. Counts only — the browser smooths the intervals.Measured directly DPI testPush along a ruler a fixed distance to estimate the real DPI, and how much further you can push.Estimate Pointer accelerationThe acceleration curve at the turn, to see whether the OS is changing your movement.Shape only Keyboard104 keys, with press counts and hold time; releases are replayed when the window loses focus.Measured directly Hold & dragA brief break looks exactly like release-and-press, so we can only say “suspected drop”.Shape only Double-click testLogs every gap, flags suspected chatter under 50 ms, and reports the median.Measured directly Wheel testDirection is reliable; step counts are estimates — the browser never says how big a step is.Direction only

Display

2 tools

Dead pixel testFive full-screen colours to check with your own eyes — it reports nothing, so it isn’t marked as a measurement.No measurement Refresh rate (visual)A moving block plus the measured rAF cadence. “It looks janky” can also come from the compositor.Estimate

Audio

2 tools

Microphone testLive waveform and level. Stated up front: the browser’s auto-gain changes the signal, so it is not SPL.Estimate Speaker / headphonesLeft/right channels, sweep, polarity and volume steps — for imbalance, distortion and noise.No measurement

Camera

1 tool

Camera testThe resolution and frame rate actually received — requested and actual are reported separately.Measured directly

Utilities

9 tools

Screen RecorderRecord the screen, microphone and a camera picture-in-picture. System audio needs the share-dialog checkbox.Not a measurement Image converterConvert between PNG / JPEG / WebP and resize. Stated clearly: re-encoding is lossy, metadata is dropped, HEIC is not supported.No measurement Favicon generatorOne image → a favicon and platform icons, each downloadable.No measurement Hash toolMD5 / SHA-1 / SHA-256. Explained properly: a hash is not encryption and cannot be reversed.No measurement CRC32 checkerCompute a checksum for text or a file, and compare it with a value you already have.No measurement JSON formatterValidate, format and minify — with the exact line and column when something is wrong.No measurement Base64 / URL encoderHandles Chinese text and binary correctly, not just a string round-trip.No measurement QR code generatorText or a link → a QR code, with size and error correction, downloadable as SVG.No measurement Unit converterLength, mass, temperature, angle and radix — conversion is not measurement, hence “No measurement”.No measurement

Practice

6 tools

Reaction timePress as soon as it lights up; the median of several rounds. It tells you the number is necessarily high.Estimate Aim LabConvert eDPI, cm/360° and FOV for Valorant / CS2, and choose a training mode. The range is still in development.No measurement AimHits per second plus average miss distance. The hit rate is always printed with the target diameter.Measured directly SteadinessThe pointer’s drift, not your hand’s — the OS has already smoothed it.Estimate Doodle boardJust for fun; it reports nothing.No measurement Sensitivity convertereDPI and cm/360°. The DPI you enter is nominal — measure the real value first.Estimate

Start with these

If you only have five minutes, start with these four.The preview images are placeholders and will be replaced before launch.

Mouse buttons / wheelA mouse diagram — press a button and that part lights up.

Dead pixelsFive solid colours, full screen.

Keyboard104 keys, none missed.

Reaction timeMedian of five rounds — no false precision.

When you’d use it

These three photos are placeholders for real shots, to be replaced with the studio’s own before launch.

A keyboard and mouse on a desk — **New keyboard or mouse**Test it out of the box, and return it within the seven-day window if something is wrong.

A dual-monitor desk with headphones — **New monitor or headphones**Dead pixels, channel imbalance, left/right — five minutes.

An office desk with peripherals — **Suspect it’s broken**Test it yourself before deciding whether to send it in.

Why every number is labelled

Some numbers on this site are measured directly, and some are only estimates or approximations: browser timer precision, mouse polling interval and monitor response time all affect the reading.

So every page labels its numbers — Measured directly / Estimate / Shape only / Direction only / No measurement — and the text explains where the number comes from and roughly how accurate it is.

A clearly labelled approximation is more useful than a precise number with no stated source.

One toolbox,for everyday checks and conversions