DeviceMetrik

Refresh Rate Test

Estimate the current browser presentation timing from requestAnimationFrame intervals.

Browser-observed measurementPrivacy-first

Tests run locally in your browser. We do not upload keystrokes, pointer paths, or media.

Important

  • This estimates browser presentation timing, not the panel’s advertised maximum refresh rate.

This is an estimate of the current browser presentation timing, not the monitor panel's advertised maximum refresh rate.

—

Estimated from median of robust-filtered intervals

Fullscreen not supportedWake Lock not supported

Results

Session
Ready
Estimated refresh rate
—
Warm-up duration
0 ms
Measurement elapsed
0 ms
Median interval
—
P95 interval / MAD
—

P95 is a slower extreme interval. MAD is median absolute deviation; lower means more consistent intervals.

Stability
—

1 − (MAD ÷ median interval); higher means more consistent timing

Raw post-warmup
0
Filtered estimate samples
0
Excluded outliers
0

Excluded only from the robust refresh estimate; raw data remains

Long frames (raw)
0
Severe stalls (raw)
0
Possible VRR variation
Not indicated
Interrupted reason
—

How to use

  1. Choose 3s, 5s, or 10s measurement window.
  2. Click Start. Warm-up samples are discarded.
  3. Keep the tab focused; avoid resizing the window.
  4. Read estimated Hz from the median filtered interval.

What the result means

  • Estimated Hz = 1000 / median frame interval (ms).
  • Filtered samples exclude severe stalls; raw count stays visible.
  • Possible VRR variation flags large interval spread.

Limitations

  • Not the panel’s advertised maximum refresh rate.
  • Browser, compositor, power saving, and VRR affect timing.
  • Values are not force-rounded to 60/120/144 marketing labels.

Troubleshooting

Run interrupts immediately
Stay focused on the tab and avoid window resizes or monitor changes mid-test.

Related tools

FAQ

Why is my result 59.94 instead of 60?

We keep observed medians. Many displays run at fractional rates; we do not invent factory labels.