1.The Arithmetic Illusion of Average FPS
Frames Per Second (FPS) is a simple count of how many images were rendered across a full 1,000 millisecond window. However, FPS is an aggregate summary that obscures uneven time distribution. If 59 frames render in 5ms each (295ms) and the 60th frame takes 705ms to render due to a background stall, the recorded frame rate is still 60 FPS, yet the user experiences a massive, jarring hitch.
2.The Millisecond Perspective: Frame Time
Frame time measures the exact duration required to compute and present each individual frame (t_frame = 1000 / FPS). At 60 FPS, the ideal frame time is 16.67ms. At 144 FPS, it is 6.94ms. At 240 FPS, it is 4.17ms. Plotting frame time as a real-time scatter graph exposes micro-stutter spikes that average counters completely conceal.
3.Understanding 1% Low and 0.1% Low Metrics
To quantify micro-stutter objectively, engineers evaluate the bottom 1st percentile and 0.1th percentile of frame times. The 1% low metric represents the average frame rate of the slowest 1% of frames during the test session. If your average FPS is 144 but your 1% low is 42 FPS, your eyes will frequently perceive stuttering.
Frequently Asked Questions
Why is frame time more important than average FPS for competitive gaming?
Consistent low frame times (e.g. 6.94ms ± 0.5ms) minimize input latency variability, allowing muscle memory and tracking precision to remain consistent without sudden frame delivery delays.
How do I test my own frame time variance online?
Use the FPSRate Frame Time Test to view your real-time frame time graph, median P50 latency, P95 tail latency, and stutter spikes directly in your browser.