Test 06
Refresh rate test
Measures the real time between frames, not what the specification claims. Then checks how evenly those frames arrive, how cleanly the panel handles motion, and how fast it reacts to a click.
Motion and latency
Two strips run at once: a dark pattern on light, and a light pattern on dark. Most panels are noticeably slower on one of the two, and running both side by side is the only way to see the difference without a camera.
Expect roughly half your frame time on a healthy setup, because a click lands on average halfway through a frame. At 60Hz that is around 8ms, at 144Hz around 3.5ms. Much higher means something is buffering, often a browser extension or a background process.
The average is the least useful number
A display can report a perfect 144Hz and still feel rough, because what you notice is not the average frame but the worst one. That is why the 1% low and the worst frame are shown separately. The 1% low is the average of the slowest one percent of frames, the same figure game benchmarks report, and it is where stutter actually lives.
If your average sits at your rated rate but the 1% low is two or three times higher, the panel is fine and something is interrupting the browser. Background processes, browser extensions, and power saving profiles are the usual causes.
Jitter and the 1% low often disagree, and that is useful rather than contradictory. Jitter is simply the largest frame minus the smallest, so a single hiccup dominates it. The 1% low ignores that. When jitter looks bad but the 1% low and consistency are clean, your display is smooth and you caught one interruption.
Monitor response time test: what a browser can and cannot do
Response time is the figure printed on the box, usually as a grey-to-grey number in milliseconds. It describes how long a pixel takes to change from one shade to another.
A browser cannot measure it. Doing so properly needs a photodiode or a high speed camera aimed at the panel, because the measurement is of light leaving the screen, not of anything the software can see. Any site handing you a grey-to-grey figure from a web page is estimating.
What a browser can do is drive the panel hard enough that you can see the result. That is what the motion test does. Follow the moving block with your eyes rather than staring at a fixed point, the way you would track a character in a game. Two faults show up clearly:
- Trailing is a smear stretching behind the block. Pixels are changing too slowly to keep up.
- Overshoot is a bright or dark halo in front of the block, sometimes called inverse ghosting. The overdrive setting is pushing pixels past their target and they are settling back.
Trailing with no halo usually means overdrive is set too low, and raising it will sharpen motion. A visible halo means it is set too high, and lowering it will clean the image up at the cost of slightly softer motion. Most monitors expose this in their menu as Overdrive, Response Time, or Trace Free.
Why two strips run at once
The motion test shows a dark pattern on white and a light pattern on black at the same time. Panels are rarely symmetrical: most VA panels are noticeably slower going from dark to light than the other way round, which is where black smearing comes from. Running both directions side by side is the only way to see that difference without instruments.
Reading the input latency number
A click lands at a random point within a frame, so on average you wait half a frame for the next one to be drawn. That puts a healthy result at roughly 8ms at 60Hz and 3.5ms at 144Hz.
This figure covers the browser and frame scheduling only. It excludes mouse polling, the operating system input stack, and the panel itself, so treat it as a way to compare setups or spot a problem rather than as a total system lag measurement. A median far above half a frame usually points at a browser extension, a heavy background process, or a power saving mode.
Common questions
How do I check my monitor refresh rate?
Click Start measuring and leave the tab in front for about five seconds. The large number is your measured refresh rate in Hz, taken from the median time between frames.
I bought a 144Hz monitor but this shows 60Hz. Why?
Displays very often default to 60Hz until you change it. On Windows, open Display settings, then Advanced display, and select the higher rate. Also check the cable, since some HDMI cables cannot carry high refresh rates at full resolution.
Can a browser measure my monitor response time in milliseconds?
No, and any site that gives you a grey-to-grey figure is estimating rather than measuring. Real response time is measured with a photodiode or a high speed camera pointed at the panel. What a browser can do is show you moving patterns so you can see trailing and overshoot for yourself, which is what the motion test here is for.
What is trailing and what is overshoot?
Trailing is a smear behind a moving object, and it means pixels are changing too slowly. Overshoot is a bright or dark halo in front of the object, and it means the overdrive setting is pushing pixels too hard. Turning overdrive down usually removes overshoot at the cost of a little more trailing.
What is the 1% low?
It is the average of the slowest one percent of frames measured, the same figure game benchmarks report. A display can average a perfect 144Hz and still feel rough, because what you notice is the worst frames rather than the average. If the 1% low sits far above your normal frame time, something is interrupting the browser.
Why does the 1% low say sampling?
It needs at least two hundred frames before one percent of them is more than a single frame. Until then the number would look precise without being meaningful, so the test waits. Leave it running for about four seconds and it fills in.
Why is the reading slightly off, like 59.9 or 143.8?
That is normal and expected. Refresh rates are rarely exact whole numbers, and browser timing carries a small amount of noise. Anything within a couple of Hz of a standard rate is your display running correctly.
What does the input latency number include?
It measures the gap between your click reaching the browser and the next frame being drawn, so it covers browser processing and frame scheduling. It does not include mouse polling, the operating system input stack, or the panel itself, so it is a useful relative figure rather than a total system lag measurement.
Why did my reading collapse when I switched windows?
Browsers throttle background tabs to around one frame per second to save power. The test detects this, pauses, and restarts the measurement when you return.