Refresh Rate and Frame Time Test
The meter samples animation callbacks and estimates their cadence, average interval, variation and long intervals. Results depend on the browser, system load and available timer precision.
Open tool fullscreen →Interface and local controls remain in Portuguese.
How to use
- Close heavy tabs, keep this tab in the foreground and plug the laptop in — power saving caps the composition cadence.
- Wait 5 to 10 seconds until the average Hz settles; the instant value naturally swings over short sampling windows.
- Compare the result with the rate configured in the operating system and watch the frame time graph for periodic spikes.
When to use
- Compare observed cadence with the configured rate after swapping cable, port or driver.
- Record cadence drops for investigating load, power, compositor or other causes.
- Observe callback regularity in different modes without validating VRR.
- Record before-and-after evidence of configuration changes in technical support reports.
Limitations
The browser cannot exceed the system compositor cadence, and it measures neither panel response time nor total input lag. The number is the composition rate as perceived by the tab.
Frequently Asked Questions
- My monitor is 165 Hz but it measured 60 Hz. Why?
- The tab may be limited by the browser, power saving, compositor or system configuration. Check the display mode and repeat in the foreground; the page cannot identify the cause.
- What does high variation with a correct average Hz mean?
- Callback intervals were irregular during the sample. System load, scheduling and composition can influence them; the value does not diagnose the panel or VRR.
Technical Methodology & Display Science
Counting relies on requestAnimationFrame, whose callback is scheduled right before each frame composition and receives a high-resolution timestamp as defined by the W3C High Resolution Time specification. The difference between consecutive timestamps estimates callback intervals; it does not confirm physical presentation time.
The average rate divides the frame count by the total elapsed interval, which removes short-window noise. Jitter is the standard deviation of frame times, computed from an accumulated sum of squares; this describes sample regularity and does not diagnose panel health.
The dropped-frame counter flags every interval above 1.8 times the running mean, a heuristic for long intervals, not a certified count of discarded frames. The measurement does not verify VRR operation.