Uneven motion, blur and ghosting can look similar at first. A marker might pause and jump, move regularly while looking blurred, or leave a bright or dark outline. Start by observing which behavior you can actually see.

Look for pauses in the interval graph

Consider two illustrative one-second samples, each containing 60 intervals. In the first, every interval is about 16.67 ms. In the second, one interval is 50 ms and the other 59 are about 16.10 ms. Both average roughly 60 callbacks per second, but the second contains a much longer gap.

This is why the FPS test includes an interval chart. Read the overall cadence, then look for tall isolated intervals or repeated patterns. Chrome’s performance tooling documentation describes a more detailed workflow for investigating browser work when a simple timing sample is not enough.

What the timing statistics mean here

  • Median interval: the central interval after sorting; for an even number of intervals, the average of the two central values. It describes a typical interval without hiding the rest of the graph.
  • 95th-percentile interval: the interval at the nearest rank covering 95% of the sorted sample. A larger value points to longer intervals in the sample’s upper end.
  • Longest interval: the largest gap observed during this run.
  • Spikes: intervals greater than 1.5 times that run’s median. This threshold highlights unusually long gaps within the current sample.

The chart shows the recorded intervals without smoothing. A single spike deserves context: background work, a window change or an interaction may have happened at the same time. Note those events when comparing runs.

Compare blur and trails in a moving pattern

A dark smear behind a moving object, or a bright halo around it, can remain visible even when browser intervals are regular. Use a visual pattern alongside the timing chart to compare the trail itself.

Use the motion comparison at one fixed speed. Follow a marker and compare its edge against light and dark backgrounds. If your monitor has overdrive or response settings, compare one setting at a time and keep the refresh mode unchanged. A setting that reduces one trail may introduce a more obvious outline.

Make the next test answer one question

  1. Run a clean sample with this page’s extra load off.
  2. Repeat once under the same conditions to see whether an unusual timing pattern recurs.
  3. For timing, change one workload condition. For trails, keep workload fixed and compare the visual pattern.
  4. Save the result and note the change, rather than combining several fixes into one unexplained improvement.

Separate frame intervals from pixel response and input delay. If your basic cadence is lower than the display mode you selected, work through the setup checklist first.

Game rendering adds another set of conditions to compare. See how to distinguish game-image trails from display behavior.