How we test
Updated
Every tool on PC Test Tools is checked by hand before it goes live, and again after major browser updates. This page lists the devices and browsers I test on, how I check each type of tool, and what a browser can't measure.
Devices I test on
This is the gear I test with. Each tool gets tried on the devices it's meant for before it goes live.
- Screens: an external monitor that runs at 144 Hz or faster, a Lenovo ThinkPad running Windows, an M1 MacBook Air and a TV.
- Phones: an iPhone 15 Pro (120 Hz) and a OnePlus Nord (90 Hz).
- Keyboards: a Keychron K8 mechanical keyboard with n-key rollover, plus the built-in keyboards of both laptops.
- Mice and clickers: a Logitech G502 with side buttons and a tilt wheel, both laptops' touchpads and a Logitech R400 presentation clicker.
- Controllers: an Xbox controller, a PlayStation controller and a third controller from another brand.
- Cameras: a Logitech C920 USB webcam (1080p, 30 fps), plus the cameras built into both laptops and both phones.
Browsers
I test in current versions of Chrome and Edge on Windows. Tools that use newer browser features also get checked in Firefox, Safari on a Mac and an iPhone, and Chrome on Android. When a browser lacks a feature, the tool shows a clear message instead of failing silently.
Automatic checks on every change
Each change to the site runs the same automatic checks before it is published:
- Unit tests for the math behind every tool, such as refresh-rate statistics, click-speed windows and report formats.
- Browser tests that load each page on a desktop and a phone-sized screen, run the main test with simulated devices and fail on any console error.
- Accessibility scans with axe, plus keyboard-only checks of the main controls.
- Page checks for titles, descriptions, headings, links and structured data.
How I check accuracy
Automatic tests prove the code does what it should. They can't prove that a result matches your hardware, so each tool that measures something also has a reference check:
| Tool | Reference | Passes when |
|---|---|---|
| Refresh rate test | Windows display settings (Settings → System → Display → Advanced display) | Within 1 Hz of the set rate on a quiet system |
| CPS test | Scripted clicks at a known rate, and my own clicks counted in a slow-motion phone video | Within 0.2 clicks per second of the reference |
| Double click test | Scripted click pairs with known gaps, and a healthy mouse clicked normally | Flags every gap under the threshold and none from normal clicking |
| N-key rollover test | The keyboard's published rollover (6KRO or NKRO) | Same number of keys over USB |
| Gamepad tester | Windows game controller settings (joy.cpl) or Steam | Same buttons, axes and deadzone behavior |
| Webcam test | The camera app on the same device | Same resolution, and frame rate within 1 fps |
What a browser can't measure
Some numbers need lab equipment. A browser can't read brightness in nits, true color accuracy, a panel's response time in milliseconds, or end-to-end input lag without a camera. Pages that touch these topics say so and point to a better method, such as a colorimeter or a slow-motion phone video.
How often tools are rechecked
Once a month I run the browser tests in Firefox and Safari's engine, then open every live tool and run its main test. Every 6 months, or right away when a browser update breaks something, I review the tool and its page in full. Each page shows the date of its last meaningful update.
Report a problem
If a tool gives a result that looks wrong, please email contact@pctesttools.com with the page address, your device, operating system and browser, and what you expected to see. I reproduce every report, fix the cause and add a test so it doesn't come back.